TOWARD	THE	DEVELOPMENT	OF	SYNTHETIC	MICROBIAL	CONSORTIA	UTILIZING	
ENGINEERED	CYANOBACTERIA	AND	HETEROTROPH	INTERACTIONS	

					
By	
Derek	T.	Fedeson	
													

A	DISSERTATION	
Submitted	to	
Michigan	State	University	
in	partial	fulfillment	of	the	requirements	
for	the	degree	of	
Genetics	–	Doctor	of	Philosophy	
2018	
	

	

	
	

	
	
	

ABSTRACT	

By	
	

Derek	T.	Fedeson	

	
TOWARD	THE	DEVELOPMENT	OF	SYNTHETIC	MICROBIAL	CONSORTIA	UTILIZING	
ENGINEERED	CYANOBACTERIA	AND	HETEROTROPH	INTERACTIONS	
	
	
Cyanobacteria	form	the	foundation	of	the	trophic	cascade	of	nearly	all	biospheres	on	
the	surface	of	our	planet	and	are	ubiquitous	to	every	known	environment.	These	unique	
and	diverse	organisms	have	recently	come	to	the	forefront	of	contemporary	discussions	on	
the	sustainable	production	of	fuels,	pharmaceuticals,	cosmetics,	and	plastics.	This	increase	
in	attention	is	due	to	many	factors	including	their	high	photosynthetic	capacity,	relatively	
minimalistic	growth	requirements,	genetic	tractability,	and	untapped	potential	for	novel	
therapeutics.	While	direct	engineering	of	cyanobacterial	species	has	been	the	primary	focus	
of	present	research	trends,	there	is	growing	interest	in	utilizing	cyanobacteria	in	the	
context	of	synthetic	microbial	consortia	to	accomplish	similar	biotechnological	goals.	This	
approach	allows	for	the	distribution	of	metabolic	load	amongst	the	different	participating	
species	and	compartmentalization	of	specialties	that	in	the	context	of	cyanobacterial	driven	
consortia	could	mean	that	the	cyanobacteria	fulfill	the	role	of	providing	nutrients	to	other	
members	of	the	consortia	in	the	form	of	fixed	carbon	or	nitrogen.	The	goal,	in	essence,	
would	be	to	create	a	synthetic	symbiotic	or	commensal/cross-feeding	interaction	between	
the	cyanobacteria	and	the	other	members	of	the	synthetic	community.	However,	there	are	
many	considerations	that	need	to	be	addressed	when	conceptualizing	how	artificial	
consortia	might	be	applied	on	an	industrial	scale	which	include	maintaining	species	
specificity	in	these	consortia,	facilitating	nutrient	exchange	between	species,	preventing	
invasion	of	contaminating	microbes,	enabling	the	use	of	contaminated	water	sources,	and	

identification	of	compatible	metabolisms.	Here,	I	present	my	work	to	address	many	of	
these	questions	through:	A)	the	development	of	a	functional	surface	display	system	that	
allows	for	mediated	binding	of	cyanobacteria	to	functionalized	cells/particles,	B)	the	
expansion	of	the	surface	display	system,	C)	the	creation	of	a	photosynthetic	co-culture	
capable	of	remediating	waste	water	contaminated	with	a	xenobiotic,	and	D)	the	
experimental	evolution	of	a	heterotrophic	species	in	long	term	co-culture	with	alginate	
bead	encapsulated	cyanobacteria.	These	approaches	were	designed	to	both	mimic	
naturally	occurring	microbial	partnerships,	which	utilize	spatial	co-localization	of	
symbiotic	species	to	ensure	partnership	stability	and	ward	off	potential	contaminants,	and	
provide	practical	utility	in	an	industrial	context.	Furthermore,	this	work	lays	the	
foundation	for	comprehensive	study	of	how	cell-cell	adhesion	and	long-term	co-evolution	
could	influence	symbiotic	relationships	between	microbial	organisms,	with	the	potential	to	
yield	insights	into	the	conditions	that	eventually	led	to	the	endosymbiotic	event	that	gave	
rise	to	the	chloroplasts	found	in	eukaryotic	photosynthetic	species.

This	dissertation	is	dedicated	to	my	family.	

gratitude.	

To	my	wife	and	my	daughter,	I	love	you	both	and	you	are	the	source	of	my	inspiration.	I	
could	not	have	done	this	without	you	and	there	are	no	words	capable	of	expressing	my	
To	my	parents,	your	support	and	unwavering	faith	in	me	have	meant	more	than	you	know.	

To	my	siblings,	thank	you	for	always	being	there	when	I	needed	you.	

	

	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	

iv	

	

	

	
	

ACKNOWLEDGMENTS	

During	my	time	here	at	the	Michigan	State	University,	I	was	fortunate	to	interact	
with	many	wonderful	and	talented	individuals	who	spurred	my	development	academically	
and	provided	the	support	necessary	to	develop	this	research.		
	
First	and	foremost	I	would	like	to	thank	my	mentor,	Dr.	Daniel	Ducat.	Without	his	
knowledge	and	guidance,	none	of	this	work	would	have	been	possible.	His	kind	and	
compassionate	nature	set	the	tone	for	my	experience	here	at	MSU	and	I	count	myself	lucky	
to	have	been	one	of	his	students.	Thank	you	all	Ducat	lab	members	for	your	camaraderie!	I	
would	also	like	to	thank	my	committee	members,	Dr.	Christoph	Benning,	Dr.	Beronda	
Montgomery,	and	Dr.	Timothy	Whitehead	for	their	expertise	and	constructive	feedback	on	
my	work	over	the	years.	Their	perspectives	and	advice	have	helped	me	develop	and	grow	
as	a	scientist.	A	final	thanks	and	acknowledgment	to	the	many	wonderful	individuals	I	have	
met	during	my	time	as	part	of	the	PRL	community	and	as	a	student	in	the	Genetics	Program	
here	at	MSU.		
There	are	no	words	to	describe	the	gratitude	I	feel	toward	family	and	friends	for	all	
	
of	their	support.	Julia,	Claire,	Jodi,	Brian,	Gwen,	Jonathan	Sr.,	Jonathan	Jr.,	Erin,	Renee,	Amy,	
Courtney,	Joyce,	Thomas,	Rose,	Ryan,	Jimmy,	Fernando,	Stephanie,	Martin,	and	Kristen,	you	
are	all	amazing	and	I	am	blessed	to	have	you	in	my	life.		
	

	

	

v	

LIST	OF	TABLES	...................................................................................................................................................	ix	

TABLE	OF	CONTENTS	

	
	

LIST	OF	FIGURES	...................................................................................................................................................	x	

CHAPTER	1:	INTRODUCTION	............................................................................................................................	1	
Abstract	...............................................................................................................................................	2	
Introduction	......................................................................................................................................	3	
Part	I:	Natural	photosynthetic	microbial	communities	of	ecological	and	
technological	relevance	.................................................................................................................	6	

Cyanobacterially-driven	marine	ecosystems	.....................................................................................	6	
Plant-cyanobacterial	symbioses	...........................................................................................................	11	
Synthetic	microbial	ecology	&	microbial	ecology	theory	...........................................................	14	
The	biotechnological	potential	of	synthetic	consortia	................................................................	18	
Synthetic	co-cultures	for	photosynthesis-driven	bioindustry	.................................................	21	
Limitations	in	synthetic	co-culture	approaches	and	future	perspectives	..........................	26	

Part	II:	Promise	and	current	limitations	of	the	application	of	synthetic	microbial	
communities	....................................................................................................................................	14	

Concluding	remarks	......................................................................................................................	29	

CHAPTER	2.1:	CYANOBACTERIAL	SURFACE	DISPLAY	SYSTEM	MEDIATES	ENGINEERED	
INTERSPECIES	AND	ABIOTIC	BINDING	......................................................................................................	30	
Abstract	.............................................................................................................................................	31	
Introduction	....................................................................................................................................	32	
Results	...............................................................................................................................................	34	

Design	and	screening	of	surface	display	epitopes	.........................................................................	34	
Extracellular	factors	occlude	outer	membrane	accessibility	in	living	cells	.......................	36	
LPS	extracellular	occlusion	of	epitope	...............................................................................................	37	
Identification	and	deletion	of	extracellular	proteins	...................................................................	38	
Elimination	of	the	putative	s-layer	and	OAg	improves	surface	display	availability	.......	39	
Surface	display	mediates	interactions	with	functionalized	abiotic	beads	..........................	40	
Surface	display	mediates	binding	to	engineered	yeast	binding	partners	...........................	43	
S.	elongatus	cell	culture	and	strains	.....................................................................................................	50	
Immunostaining	of	cyanobacterial	cells	............................................................................................	50	
Cyanobacteria-bead	binding	assays	....................................................................................................	51	
Yeast	culturing	and	induction	conditions	.........................................................................................	52	
Cyanobacteria-yeast	binding	assays	...................................................................................................	52	
Surface	component	stripping	.................................................................................................................	52	
Western	blotting	..........................................................................................................................................	53	
Flow	cytometry	analysis	...........................................................................................................................	53	
	

Discussion	........................................................................................................................................	45	
Materials	and	Methods	................................................................................................................	50	

APPENDIX	............................................................................................................................................................	55	

	

vi	

CHAPTER	2.2:	ENGINEERING	OF	PAIRED	CYANOBACTERIAL	AND	HETEROTROPHIC	SURFACE	
DISPLAY	FOR	INTERSPECIES	ADHESION	...................................................................................................	76	
Introduction	....................................................................................................................................	77	
Results	...............................................................................................................................................	80	

Design	of	surface	display	strains	..........................................................................................................	80	
Evaluating	the	ability	of	the	surface	displayed	moieties	with	soluble	conjugate	proteins
	.............................................................................................................................................................................	86	
Strains	and	culture	conditions	...............................................................................................................	91	
Construct/strain	development	..............................................................................................................	91	
Immunostaining	and	western	blotting	...............................................................................................	93	
Protein	reagent	purification	...................................................................................................................	93	

Discussion	........................................................................................................................................	89	
Materials	and	Methods	................................................................................................................	91	

CHAPTER	3:	BIOTRANSFORMATION	OF	2,4-DINITROTOLUENE	IN	A	PHOTOTROPHIC	CO-
CULTURE	OF	ENGINEERED	SYNECHOCOCCUS	ELONGATUS	AND	PSEUDOMONAS	PUTIDA	..........	95	
Abstract	.............................................................................................................................................	96	
Importance	......................................................................................................................................	97	
Introduction	....................................................................................................................................	98	
Results	............................................................................................................................................	103	

Alginate-encapsulated	S.	elongatus	CscB	can	tolerate	2,4-DNT	at	higher	concentrations	
than	planktonic	cultures	.........................................................................................................................	103	
Engineering	P.	putida	EM173	for	sucrose	consumption	and	evaluation	of	growth	
parameters	in	the	presence	of	alginate-encapsulated	S.	elongatus	CscB	..........................	106	
Growth	of	engineered	P.	putida	strains	is	supported	by	sucrose-rich	cyanobacterial	
exudates	in	a	synthetic	consortium	system	...................................................................................	109	
Biotransformation	of	2,4-DNT	by	engineered	P.	putida	EM·DNT·S	in	both	monoculture	
and	co-culture	.............................................................................................................................................	110	
Degradation	of	2,4-DNT	by	a	synthetic	consortium	of	P.	putida	EM·DNT·S	and	alginate-
encapsulated	S.	elongatus	CscB	and	long-term	culture	potential	.........................................	115	
Simultaneous	2,4-DNT	biodegradation	and	PHA	bioproduction	by	engineered	strains	
in	a	synthetic	consortium.	......................................................................................................................	116	
Bacterial	strains	and	culture	conditions..........................................................................................	122	
Encapsulation	of	S.	elongatus	CscB	in	alginate	beads	................................................................	123	
Analytical	methods	...................................................................................................................................	124	
Selection	and	engineering	of	the	ancestral	strain	.......................................................................	140	
Experimental	selection	conditions	for	photosynthetic	co-culture	.......................................	143	
Long-term	co-culture	to	select	for	enhanced	fitness	.................................................................	146	

CHAPTER	4:	DIRECTED	EVOLUTION	OF	ESCHERICHIA	COLI	W	CSCR	WITH	ALGINATE	
ENCAPSULATED	SYNECHOCOCCUS	ELONGATUS	CSCB	..........................................................................	136	
Introduction	.................................................................................................................................	137	
Results	............................................................................................................................................	140	

Discussion	.....................................................................................................................................	117	
Methods	..........................................................................................................................................	122	

APPENDIX	..........................................................................................................................................................	127	

	

vii	

Preliminary	verification	of	enhanced	fitness	in	evolved	lines	...............................................	147	
Bacterial	strains,	media	and	growth	conditions	..........................................................................	152	
Co-culture	conditions	...............................................................................................................................	152	
Imaging	and	spectrophotometery	......................................................................................................	153	

Discussion	.....................................................................................................................................	149	
Materials	and	Methods	.............................................................................................................	152	

CHAPTER	5:	CONCLUSIONS	AND	FUTURE	PROSPECTIVES	................................................................	154	
Overview	........................................................................................................................................	155	
Diffuse	resources	and	spatial	structure	..............................................................................	156	
Natural	examples	of	structured	consortia	.........................................................................	157	
Artificial	consortia	and	modularity	......................................................................................	159	
Alginate	encapsulation	as	a	technique	for	artificial	consortia	...................................	160	
The	intersection	of	synthetic	ecology	and	experimental	evolution	..........................	161	

REFERENCES	.....................................................................................................................................................	163	

		

	

	

viii	

LIST	OF	TABLES	

Table	2.1.S1:	Complete	LC-MS-MS	peptide	reads	and	predicted	protein	matches	for	

	
EDTA-solubilized	proteins	..............................................................................................................	62	
Table	2.1.S2:	Primer	List	.................................................................................................................	72	
Table	2.2.1:	Chapter	2.2	Primer	list	.............................................................................................	92	
Table	3.S1:	Media	composition	..................................................................................................	128	

	

	

	

	

	

	

ix	

	

LIST	OF	FIGURES	

Figure	1.1:	Study	of	microbial	ecology	at	different	levels	of	abstraction	.......................	15	
	
Figure	1.2:	Modularity	in	microbial	consortia	.........................................................................	22	
	
Figure	2.1.1:	Graphical	abstract	....................................................................................................	34	
	
Figure	2.1.2:	SomA	surface	display	design	and	efficacy	........................................................	35	
	
Figure	2.1.3:	Surface	display	epitope	is	occluded	by	EDTA-sensitive	extracellular	
factor(s)	.................................................................................................................................................	37	
	
Figure	2.1.4:	Removal	of	OAg	synthesis	machinery	and	a	putative	S-layer	protein	
improves	epitope	availability	........................................................................................................	38	
	
Figure	2.1.5:	Surface	display	can	mediate	interactions	between	cyanobacteria	and	
abiotic	surfaces	...................................................................................................................................	42	
	
Figure	2.1.6:	Surface	display	mediated	adhesion	of	cyanobacteria	to	engineered	
yeast	........................................................................................................................................................	45	
	
Figure	2.1.S1:	Comparisons	of	SomA	models	and	NS3	tagged	SomA	insertion	
construct	design	.................................................................................................................................	58	
	
Figure	2.1.S2:	Additional	representative	immunolocalization	of	indicated	tagged	
SomA	constructs	in	fixed	cells	.......................................................................................................	59	
	
Figure	2.1.S3:	Cyanobacterial	cell	viability	following	EDTA	treatment	..........................	60	
	
Figure	2.1.S4:	O-antigen	knockout	construct	design	and	cell	morphology	of	resultant	
knockout	strains	................................................................................................................................	61	
	
Figure	2.1.S5:	Potential	surface	layer	protein	vector	constructs	and	morphology	.....	67	
	
Figure	2.1.S6:	Additional	representative	images	of	live	R5F	wzt	slpA	cells	...................	68	
	
Figure	2.1.S7:	Longer	induction	of	SomA-R5F	increases	epitope	availability	...............	69	
	
Figure	2.1.S8:	Association	of	R5FEDTA	S.	elongatus	cells	with	magnetic	Protein	A	beads	
is	dependent	upon	IPTG-induced	expression	of	SomA-R5F	and	mediating	antibodies
	..................................................................................................................................................................	70	
	

x	

	

Figure	2.1.S9:	Induction	of	Protein	A	on	the	surface	of	EY100	S.	cerevisiae	and	
antibody-mediated	adhesion	of	yeast	cells	to	R5F	S.	elongatus	.........................................	71	
	
Figure	2.2.1:	S.	elongatus	SpyTag	and	Strep-tag	II	surface	display	constructs	..............	82	
	
Figure	2.2.2:	S.	cerevisiae	SpyCatcher	and	StrepCoreMut2	surface	display	constructs
	..................................................................................................................................................................	84	
	
Figure	2.2.3:	E.	coli	W	cscR	SpyCatcher	surface	display	construct	....................................	86	
	
Figure	2.2.4:	Protein	reagent	development	..............................................................................	87	
	
Figure	2.2.5:	Preliminary	testing	of	S.	cerevisiae	and	S.	elongatus	strains	with	protein	
reagents	.................................................................................................................................................	89	
	
Figure	3.1:	Conceptual	design	of	the	photosynthetic	co-culture	designed	for	
simultaneous	biodegradation	and	bioproduction	...............................................................	101	
	
Figure	3.2:	Growth	and	physiological	parameters	of	planktonic	and	alginate-
encapsulated	S.	elongatus	CscB	..................................................................................................	104	
	
Figure	3.3:	Construction	of	P.	putida	EM·DNT·S	and	characterization	of	sucrose-
dependent	growth	alone	or	in	co-cultures	.............................................................................	107	
	
Figure	3.4:	2,4-DNT	biotransformation	in	monocultures	of	engineered	P.	putida	...	112	
	
Figure	3.5:	Degradation	of	2,4-DNT	in	co-culture	and	long-term	co-culture	cycling	114	
	
Figure	3.6:	PHA	accumulation	in	P.	putida	bioremediating	co-cultures	......................	117	
	
Figure	3.S1:	Influence	of	2,4-DNT	on	the	growth	of	S.	elongatus	PCC	7942	.................	129	
	
Figure	3.S2:	Encapsulated	S.	elongatus	CscB	in	different	media	with	2,4-DNT	..........	130	
	
Figure	3.S3:	Calculated	chlorophyll	a	per	cell	comparison	between	planktonic	and	
encapsulated	S.	elongatus	CscB	..................................................................................................	131	
	
Figure	3.S4:	Comparing	P.	putida	supernatant	spectra	over	time	..................................	132	
	
Figure	3.S5:	4M5NC	control	elution	profile	and	m/z	..........................................................	133	
	
Figure	3.S6:	Loss	of	2,4-DNT	from	P.	putida	monocultures	and	reductive	pathway	
analysis	...............................................................................................................................................	134	
	
Figure	4.1	E.	coli	W	cscR	lacZ::Fluorophore	lines.	................................................................	142	
	

xi	

	

Figure	4.2:	E.	coli	W	cscR	lacZ::mNeonGreen	experimental	evolution	with	alginate	
encapsulated	S.	elongatus	CscB	..................................................................................................	144	
	
Figure	4.3:	E.	coli	W	cscR	lacZ::mNeonGreen	experimental	evolution	with	daily	
dilutions	.............................................................................................................................................	147	
	
Figure	4.4:	Preliminary	evidence	for	enhanced	fitness	of	evolved	strains	in	low	
sucrose	medium	..............................................................................................................................	148	

	

xii	

CHAPTER	1:	INTRODUCTION	

 “SYMBIOTIC	INTERACTIONS	OF	PHOTOTROPHIC	MICROBES:	ENGINEERING	SYNTHETIC	
	

This	work	has	been	submitted	and	is	currently	under	review.	

CONSORTIA	FOR	BIOTECHNOLOGY”	
Derek	T.	Fedeson	&	Daniel	C.	Ducat	
Author	Contributions:	

	

DTF	and	DCD	wrote	and	edited	the	manuscript.	
	

	

1	

	

	

	

Abstract	

	

Natural	microbial	communities	consist	of	assemblies	of	species	possessing	distinct	
metabolic	capacities.	Diversification	within	the	consortia	leads	to	division	of	labor	between	
species,	whereby	the	global	population	exhibits	functional	capabilities	that	are	possessed	
by	only	a	fraction	of	its	members.	Furthermore,	community	diversity	is	also	associated	with	
higher	bioproductivities	and	robustness	compared	to	microbial	‘monocultures.’	In	this	
review,	we	highlight	both	natural	and	engineered	interactions	between	photosynthetic	
microbes	and	other	organisms,	with	an	emphasis	on	learning	design	principles	of	microbial	
communities	through	the	process	of	building	them	from	the	“bottom	up.”	Rational	design	of	
improve	 our	
relatively	 simple	 microbial	 communities	
understanding	of	much	more	complex	natural	consortia	that	have	important	ecological	
significance.	Furthermore,	a	deeper	understanding	of	effective	design	principles	of	
microbial	communities	could	enable	the	application	of	light-driven	microbial	cultures	for	a	
variety	of	environmental	and	biotechnological	goals.	
	

likely	 to	 substantially	

is	

	

	

2	

	

Introduction	

Microbial	communities	comprise	a	substantial	proportion	of	the	biomass	on	Earth	
(Bar-On	et	al.	2018)	and	underlie	the	health	and	functioning	of	many	different	ecosystems.	
Complex	microbial	communities	contribute	to	turnover	of	a	number	of	critical	global	
biogeochemical	cycles,	including	nitrogen,	oxygen,	carbon,	sulphur,	and	phosphorous	(De	
Roy	et	al.	2014).	Furthermore,	the	composition	of	local	microbiomes	is	increasingly	
recognized	to	have	direct	and	substantial	impacts	on	the	health	of	multicellular	plants	and	
animals	(Mueller	and	Sachs	2015).	Indeed,	the	vast	majority	of	microorganisms	live	within	
multi-species	communities,	yet	we	have	relatively	limited	knowledge	about	microbial	
interactions	within	these	networks	and	how	these	interactions	shape	community	
properties	and	ecosystem	functions	(Chodkowski	and	Shade	2017).	The	majority	of	
microbiology	research	conducted	in	the	20th	century	focused	upon	axenic	(single-species)	
cultures	(Jessup	et	al.	2005).	Reductionist	microbiology	has	led	to	a	number	of	scientific	
breakthroughs	and	valuable	outcomes,	but	has	also	necessarily	isolated	microorganisms	
from	their	natural	context,	instead	emphasizing	their	behaviors	within	a	test	tube	(Little	et	
al.	2008).	Within	the	past	two	decades,	powerful	new	methodologies	have	emerged	that	
facilitate	systems-level	approaches	to	study	microbial	community	ecology.	These	methods	
include	improved	genomic	and	transcriptomic	sequencing,	mass	spectrometry-based	
metabolomics,	and	proteomics	(Franzosa	et	al.	2015).	The	development	of	such	
technologies	has	paved	the	way	to	begin	to	approach	some	of	the	most	ecologically	and	
bioindustrially-relevant	microbial	communities	that	are	involved	in	everything	from	
human	health	to	agriculture	(Tringe	and	Rubin	2005).	Yet,	while	these	technologies	have	
	

3	

	

greatly	expanded	our	capacity	to	inventory	the	total	species	and	reactions	within	microbial	
consortia,	the	complex	datasets	they	generate	do	not	illuminate	the	structure	of	these	
communities.	These	deficiencies	highlight	the	need	to	distill	generalizable	principles	that	
describe	fundamental	organization	of	disparate	communities.	Identification	of	common	
themes	of	microbial	interaction	that	translate	across	numerous	consortia	is	needed	in	
order	to	more	fully	understand	consortia	behaviors.	Furthermore,	this	knowledge	can	
provide	the	basis	for	design	principles	that	may	inform	the	engineering	of	artificial	multi-
species	consortia.	The	ability	to	customize	microbial	communities	or	redesign	existing	
microbiomes	represents	a	new	horizon	in	medicine,	agriculture,	and	bioindustry	(Costello	
et	al.	2012;	Rollié	et	al.	2012;	Gopal	et	al.	2013;	Song	et	al.	2014b;	Lindemann	et	al.	2016).				
In	this	review,	we	summarize	different	approaches	that	have	been	used	to	study	
microbial	communities	and	inter-species	interactions.	To	productively	constrain	our	
discussion	within	this	broad	field,	we	will	focus	upon	natural	communities	dominated	by	
phototrophic	 microbes	 (especially	 cyanobacteria),	 and	 emphasize	 biotechnological	
applications	relying	on	these	microbes.	We	first	briefly	summarize	a	couple	representative	
natural	photosynthetic	microbial	consortia	of	ecological	significance.	This	includes	
interactions	between	photosynthetic	microbes	within	the	open	ocean	that	contribute	a	
large	proportion	to	global	biogeochemical	cycles,	and	plant-cyanobacterial	symbioses	that	
can	influence	agricultural	productivity.	A	discussion	of	these	natural	communities	will	also	
highlight	features	evident	within	natural	microbial	consortia	(e.g.,	high	productivity,	
robustness	to	environmental	perturbations,	resistance	to	invasive	species)	that	are	lacking	
in	many	current	bioindustrial	technologies.	Yet,	the	difficulty	of	dissecting	the	interaction	
networks	within	these	natural	communities	also	illustrates	current	limitations	in	our	
	

4	

	

ability	to	uncover	fundamental	design	principles	that	underlie	desirable	traits	within	these	
communities.	
As	an	alternative	approach	for	understanding	microbial	consortia,	we	will	then	
review	the	emerging	field	of	synthetic	microbial	ecology,	which	advocates	the	use	of	a	
“bottom-up”	approach	for	understanding	microbial	consortia.	Synthetic	microbial	ecology	
is	a	term	that	broadly	describes	all	rationally	designed	ecosystems	created	by	assembling	
two	or	more	defined	microbial	populations	in	a	well-characterized	and	controlled	
environment.	By	organizing	a	relatively	small	number	of	defined	microbes	into	a	
consortium,	synthetic	ecology	allows	for	the	creation	of	greatly	simplified	interaction	
networks	relative	to	the	highly	complex	and	integrated	interactomes	of	natural	
communities	(Jessup	et	al.	2004).	Furthermore,	because	the	communities	are	built	from	the	
bottom	up,	member	species	with	established	molecular	toolkits	can	be	selected.	This	
enables	the	use	of	genetic	manipulation	to	systematically	dissect	the	functions	of	a	given	
microbe	within	a	larger	community.	In	this	way,	synthetic	microbial	ecologists	are	
attempting	to	distill	complex	interaction	networks	into	broadly-applicable	theoretic	
principles	useful	for	constructing	predictive	models	for	microbial	communities	(Prosser	et	
al.	2007).	We	will	discuss	some	applications	for	synthetic	consortia,	summarize	their	
current	limitations,	and	provide	perspectives	on	the	progression	of	this	field.		

	

	

5	

	

relevance	

Part	I:	Natural	photosynthetic	microbial	communities	of	ecological	and	technological	

Photosynthetic	microbes	are	found	in	many	symbiotic	interactions	across	most	
ecosystems,	and	cyanobacteria	appear	especially	prolific	in	their	capacity	to	form	symbiotic	
associations	with	a	wide	variety	of	both	prokaryotic	and	eukaryotic	partners.	Indeed,	
cyanobacteria	have	well-documented	mutualisms	across	many	kingdoms,	including	with	
plants,	mosses,	fungi,	sponges,	dinoflagellates,	diatoms,	and	other	bacteria	(Adams	2000;	
Usher	et	al.	2007).	Such	symbiotic	interactions	can	be	with	a	single	other	species,	or,	as	is	
more	common	in	most	environments,	a	collective	of	other	organisms.	Cyanobacterial	
symbiotic	interactions	can	be	largely	categorized	around	their	capacity	to	fix	atmospheric	
carbon	(CO2)	or	nitrogen	(N2)	and	provide	them	in	bioavailable	forms	to	associated	species	
within	their	communities.	Below,	we	discuss	two	representative	symbiotic	relationships	
involving	cyanobacteria	that	revolve	around	their	capacity	to	either	provide	fixed	carbon,	
or	fixed	nitrogen	for	partner	species.	

Cyanobacterially-driven	marine	ecosystems	

Marine	microbial	communities	are	responsible	for	as	much	as	half	of	the	global	
cycling	of	carbon,	nitrogen,	sulfur,	phosphorus,	as	well	as	many	important	micronutrients	
(Fuhrman	et	al.	2015).	Such	communities	can	be	composed	of	bacteria,	archaea,	protists,	
fungi,	and	their	respective	viruses,	which	form	the	foundation	of	the	food	webs	comprising	
larger	marine	lifeforms	(Sherr	and	Sherr	2002).	In	many	marine	environments,	primary	
production	is	mainly	attributable	to	prokaryotic	phototrophs,	which	in	turn	are	dominated	
by	the	large	cyanobacterial	groups,	 Synechococcus	and	 Prochlorococcus.	In	much	of	the	
	

6	

	

surface	waters	of	the	open	ocean,	free-living	 Prochlorococcus	are	the	most	abundant	
organisms	in	both	number	and	total	biomass.	Due	to	the	large	area	of	this	environment,	
these	cyanobacteria	are	estimated	to	be	the	most	abundant	photosynthetic	cell	type	on	
Earth	(Partensky	et	al.	1999).	As	such,	it	has	been	estimated	that	Prochlorococcus	accounts	
for	~4	gigatons	of	fixed	carbon	annually,	a	number	equivalent	to	the	primary	productivity	
of	all	croplands	(Biller	et	al.	2015).	
Prochlorococcus	is	a	broad	bacterial	group	that	descends	from	the	marine	lineage	of	
Synechococcus	and	is	classified	partially	by	some	unique	features	that	distinguish	it	from	
other	cyanobacteria.	 Prochlorococcus	is	unusually	small	for	a	photosynthetic	prokaryote	
(typically	<1	um	in	length	and	width	(Morel	et	al.	1993)),	potentially	placing	it	near	the	
physical	lower	limit	of	cell	size	for	an	oxygenic	phototroph	(Raven	1994).	Furthermore,	
Prochlorococcus	has	a	unique	set	of	divinyl	chlorophyll	derivatives	which	allow	use	of	
straightforward	spectroscopic	methods	to	quantify	the	abundance	of	these	cyanobacteria	
in	mixed	communities	(Morel	et	al.	1993).		Prochlorococcus	can	be	divided	into	clades	or	
ecotypes	that	stratify	in	the	water	column	(Biller	et	al.	2015;	Johnson	et	al.	2017)	and	are	
classified	based	on	their	adaptation	to	high-light	(HL)	or	low-light	(LL)	conditions.	Within	
the	HL	group,	the	most	recently	diverged	members	are	found	nearest	to	oligotrophic	
surface	waters	where	light	is	abundant,	but	other	essential	nutrients	are	growth	limiting	
(Braakman	et	al.	2017).	Indeed,	 Prochlorococcus	appears	to	have	experienced	strong	
selected	pressure	for	survival	in	low	nutrient	conditions	(Dufresne	et	al.	2008)	and	is	well-
adapted	for	growth	at	very	low	ambient	concentrations	of	nitrogen,	phosphate,	and	iron.	
Given	Prochlorococcus’	relatively	slow	growth	rates	and	nutrient-poor	environment,	
it	is	somewhat	surprising	that	these	cyanobacteria	have	a	high	rate	of	CO2	fixation	relative	
	

7	

	

to	other	cyanobacterial	species	(Hartmann	et	al.	2014).	This	appears	to	be	due	in	part	to	
their	secretion	of	a	large	amount	of	soluble	carbohydrates	(estimated	to	account	for	up	to	
~25%	of	total	fixed	carbon;	Bertilsson	et	al.,	2005).	Indeed,	a	recent	paper	posits	that	
under	the	nutrient-limited	environment	of	the	open	ocean,	 Prochlorococcus	has	been	
selected	to	minimize	nutrient	requirements	and	to	increase	sugar	secretion	as	a	metabolic	
sink	to	dissipate	excess	ATP/NADPH	(Braakman	et	al.	2017).	This	secretion	of	
carbohydrates	may	serve	a	second	role	in	promoting	the	formation	of	symbiotic	
interactions	with	neighboring	species,	as	it	allows	 Prochlorococcus	to	directly	provide	
carbon	(e.g.,	via	secretion)	in	addition	to	indirect	(e.g.,	lysis)	routes.	
A	number	of	heterotrophic	microbes	co-exist	in	mutualistic	relationships	with	
Prochlorococcus,	
including	 some	 of	 the	 most	 abundant	 marine	 bacteria,	 the	 α-
proteobacteria	group	SAR11	(Pelagibacterales;	Giovannoni,	2005).	Among	other	traits,	
SAR11	is	classified	by	the	lack	of	complete	glycine	and	serine	synthesis	pathways,	
deficiencies	that	can	be	compensated	for	by	supplementation	with	the	metabolic	precursor	
glycolate	(Carini	et	al.	2013).	Prochlorococcus	secretes	glycolate,	pyruvate,	citrate	and	other	
organic	acids	in	large	amounts	(Bertilsson	et	al.	2005).	These	are	dominant	sources	of	
carbon	that	promote	the	growth	of	SAR11	(Giovannoni	2017).	In	return,	Prochlorococcus	
benefits	from	SAR11	and	other	heterotrophic	“helper	bacteria”	that	can	detoxify	reactive	
oxygen	species	(ROS;	Morris	et	al.,	2011;	Zinser,	2018).	Many	SAR11	ecotypes	retain	
catalases	that	can	mitigate	hydrogen	peroxide,	while	Prochlorococcus	has	lost	the	gene	for	
this	enzyme.	Other	heterotrophic	bacteria	isolated	with	 Prochlorococcus	have	also	been	
shown	to	greatly	improve	the	persistence	of	 Prochlorococcus	cells	during	periods	of	
extended	dark	(Biller	et	al.	2016),	perhaps	by	cross	feeding	metabolites	(such	as	malate	
	

8	

the	

Despite	

(Braakman	et	al.	2017))	or	by	helping	to	maintain	synchrony	in	the	expression	of	circadian-
controlled	genes	(Biller	et	al.	2018).	These	mutualistic	interactions	may	be	especially	
important	to	Prochlorococcus	cells	that	are	carried	into	deep,	light-limited	regions	of	the	
water	column	by	internal	waves	and	oceanic	currents	(Biller	et	al.	2015).	Overall,	it	is	
proposed	that	such	mutualistic	exchanges	can	greatly	increase	the	total	biomass	
productivity	of	marine	ecosystems,	with	Prochlorococcus-derived	carbon	contributing	up	to	
40%	of	total	bacterial	production	(Bertilsson	et	al.	2005).	Given	the	huge	surface	area	of	
the	ocean,	mutualisms	that	increase	carbon	capture	efficiency	of	Prochlorococcus	have	the	
potential	to	greatly	impact	global	carbon	cycling	and	marine	ecosystem	food	webs	(Azam	
and	Malfatti	2007;	Brussaard	et	al.	2016;	Braakman	et	al.	2017).			
importance	 of	 Prochlorococcus-heterotrophic	
	
interactions,	 our	
understanding	of	even	the	best-studied	relationships	remains	relatively	limited.	There	are	
several	complications	related	to	the	analysis	of	 Prochlorococcus-SAR11	interactions	that	
illustrate	the	difficulties	encountered	in	research	of	natural	microbial	communities	in	
general.	First,	despite	considerable	diversity	within	Prochlorococcus,	individual	isolates	do	
not	differ	in	their	16S	rRNA	sequences:	a	standard	metric	for	defining	bacterial	species.	
Instead,	 Prochlorococcus	isolates	are	grouped	by	ecotype	clades,	and	though	individual	
members	may	share	a	
‘core-genome’	of	~1000	genes	that	are	shared	across	all	
Prochlorococcus,	the	“pan-genome”	is	much	larger,	with	tens	of	thousands	of	genes	
predicted	(Baumdicker	et	al.	2012;	Biller	et	al.	2015).		Many	genes	cluster	in	hypervariable	
regions	of	the	genome	that	can	differ	largely	between	otherwise	closely	related	strains:	
indeed	a	recent	analysis	of	 Prochlorococcus	populations	within	the	same	milliliter	water	
sample	have	shown	that	there	can	be	hundreds	of	distinct	co-existing	subpopulations	that	
	

	

9	

	

differ	substantially	in	genomic	content	but	which	are	stably	maintained	(Kashtan	et	al.	
2014).	Similarly,	SAR11	is	a	broad	class	of	α-proteobacteria	with	equally	difficult	
“boundaries,”	 and	 genomic	 diversity	 (Giovannoni	 2017).	
nomenclature,	 species	
Furthermore,	relatively	few	Prochlorococcus	or	SAR11	isolates	have	been	isolated	in	axenic	
cultures,	and	the	genetic	tools	available	for	these	isolates	are	quite	limited.	The	physical	
structuring	between	 Prochlorococcus	and	helper	bacteria	remains	unclear,	and	such	
physical	associations	are	known	to	strongly	influence	interaction	dynamics	(De	Roy	et	al.,	
2014;	Jessup	et	al.,	2004;	Said	and	Or,	2017;	see	below).		
Finally,	as	metagenomic	approaches	to	study	natural	environments	become	
increasingly	advanced	and	exhaustive	(e.g.,	Tara	Oceans	Expedition;	Bork	et	al.,	2015),	it	
becomes	ever	more	clear	that	the	interactions	between	Prochlorococcus	and	SAR11	are	but	
a	small	fraction	of	the	total	autotroph/heterotroph	“interactome”	in	any	given	marine	
ecosystem.	Indeed,	pairwise	studies	between	 Prochlorococcus	and	more	than	300	
heterotrophic	bacteria	isolated	from	marine	environments	indicated	that	the	majority	of	
these	 heterotrophs	 positively	
influenced	 cyanobacterial	 growth,	 suggesting	
that	
mutualisms	may	dominate	autotroph/heterotroph	relationships	in	these	environments	
(Sher	et	al.	2011).		Prochlorococcus-SAR11	interactions	may	occur	against	a	background	of	
other	oceanic	microbes,	yet	some	estimates	suggest	that	as	few	as	0.01-0.1%	of	these	
marine	species	can	be	cultured	in	the	lab	with	conventional	approaches	(Connon	and	
Giovannoni	2002).	These	complex	networks	defy	current	metagenomics	and	bioinformatics	
approaches	to	disentangle	and	assign	roles	to	individual	members	within	the	community	
(Temperton	 and	 Giovannoni	 2012;	 Zengler	 and	 Palsson	 2012),	 demanding	 the	
	

10	

Plant-cyanobacterial	symbioses	

development	of	new	approaches	to	understand	such	communities	at	a	systems	level	
(Kazamia	et	al.	2016).	
	
Cyanobacteria	have	a	rich	history	of	symbiotic	interactions	with	many	plants,	and	
are	particularly	notable	for	their	mutualistic	relationships	with	an	evolutionarily	wide	
range	of	plants	in	the	green	lineage,	from	Bryophytes	(e.g.,	hornworts,	mosses),	to	ferns,	to	
more	recently	evolved	angiosperms	(Usher	et	al.	2007;	Bergman	et	al.	2008).	Many	of	the	
most	tightly	integrated	symbiotic	interactions	are	formed	between	the	widespread,	
terrestrial	cyanobacterial	genera,	 Nostoc	and	 Anabaena.	Species	within	these	genera	are	
often	capable	of	nitrogen	fixation	and	form	differentiated	cellular	structures	called	
heterocysts	that	protect	nitrogenases	from	inactivation	by	oxygen	(Zhang	et	al.	2006).	
These	traits	are	important	because	the	capacity	to	provide	a	source	of	fixed	nitrogen	is	a	
core	feature	of	most	cyanobacterial	symbioses	with	plants.	In	addition	to	providing	
nitrogen,	cyanobacteria	can	perform	a	number	of	other	functions	that	promote	plant	health	
and	productivity,	including	secretion	of	antibiotics	that	discourage	plant	pathogens	and	
improvement	of	soil	fertility	(Dodds	et	al.	1995;	Adams	2000).	As	plant-bacterial	
interactions	are	a	focal	topic	beyond	the	context	of	this	dissertation,	we	do	not	focus	upon	
this	topic	at	length	here,	(interested	readers	are	directed	to	excellent	reviews;	Adams,	
2000;	Adams	et	al.,	2013;	Bergman	et	al.,	2008;	Rai	et	al.,	2000),	but	instead	provide	a	brief	
discussion	in	the	broader	context	of	the	study	of	microbial	communities.		
Unlike	the	symbiotic	interactions	between	cyanobacteria	and	heterotrophic	bacteria	
within	marine	habitats,	some	of	the	best	studied	examples	of	plant-cyanobacterial	
	

	

11	

	

interactions	have	a	high	degree	of	structural	definition.	Such	mutualistic	interactions	are	
often	initiated	by	the	release	of	diffusible	signals	(e.g.,	hormogonia	inducing	factor,	HIF)	
these	 signals	 stimulate	 differentiation	 of	 nearby	
from	 nitrogen-starved	 plants,	
cyanobacteria	into	infective	stages	(i.e.,	hormogonia;	Meeks	and	Elhai,	2002).	Hormogonia	
development	involves	structural	changes	and	the	expression	of	motility	genes	that	allow	
cyanobacterial	filaments	to	migrate	towards	the	plant,	often	leading	to	invasion	and	
colonization	of	predefined	plant	cavities,	open	stomata,	or	uptake	as	intracellular	
endophytes	(Rai	et	al.	2002).	Following	colonization,	inter-species	signaling	pathways	
direct	further	differentiation	of	cyanobacteria,	often	leading	to	enhanced	heterocyst	
formation	and	improved	nitrogen	fixation	capabilities.	In	this	way,	the	cyanobiont	becomes	
capable	of	increased	secretion	of	ammonium	or	other	nitrogen-containing	compounds	for	
the	benefit	of	the	host	plant.	“Loose	associations”	between	plants	and	cyanobacteria	are	
also	widespread,	where	cyanobacteria	play	important	roles	in	the	broader	microbial	
community	in	rhizosphere	surrounding	plant	roots,	or	where	cyanobacteria	grow	
epiphytically	on	plant	leaves	or	other	surfaces.			
In	return,	the	plant	provides	several	benefits	for	cyanobacterial	partners.	In	most	
cases,	the	plant	host	can	become	the	primary	source	of	carbon	for	the	cyanobiont,	secreting	
carbohydrates	into	a	localized	space,	or	by	general	secretion	of	diffusible	organic	carbon	
compounds	in	the	broader	vicinity	of	plant	structures.	For	example,	it	is	estimated	that	
many	plants	secrete	a	substantial	proportion	(up	to	20%)	of	the	total	carbon	they	fix	
through	the	roots,	where	it	can	support	the	growth	of	nearby	microbes	(Bais	et	al.	2006).	
Cyanobacteria	tightly	associated	with	plants	can	gain	other	advantages	due	to	the	
environment	that	is	provisioned	by	the	plant,	including	supply	of	additional	nutrients	(Rai	
	

12	

	

et	al.	2000),	and	protection	from	external	environmental	stresses,	such	as	desiccation	
(Adams	2000).	For	example,	in	one	of	the	better	studied	interactions	between	the	
cyanobacteria	Nostoc	azollae	and	the	free-floating	aquatic	fern	Azolla,	the	cyanobacterium	
is	housed	in	a	dorsal	leaf	cavity	where	it	is	fed	carbohydrates	and	other	nutrients.	Nostoc	
azollae	in	turn	provides	the	host	plant	with	sufficient	nitrogen	to	promote	its	rapid	growth	
even	in	relatively	nitrogen	depleted	waters	(Adams	et	al.	2013).	This	symbiotic	interaction	
is	of	major	agronomic	significance,	particularly	in	the	cultivation	of	rice,	as	Azolla	is	widely	
used	as	a	traditional	‘green	manure’	for	fertilization	of	crop	species	(Vaishampayan	et	al.	
2001).	Despite	a	more	detailed	understanding	of	some	molecular	mechanisms	of	
cyanobacteria-plant	interactions,	these	relationships	remain	difficult	to	study	and	
relatively	poorly	understood.	In	many	cases,	the	co-evolution	of	the	cyanobiont	and	host	
has	been	extensive,	leading	to	the	development	of	complex	networks	of	metabolic	exchange	
and	signaling	molecules.	Many	plant	derived	HIF	factors	are	unknown,	while	cyanobacteria	
also	employ	a	range	of	anti-hormongonia	factors	(Liaimer	et	al.	2015)	that	also	are	
relatively	poorly	understood.	Once	in	an	established	mutualistic	interaction,	the	network	of	
signals	exchanged	is	likely	extensive,	but	largely	uncharacterized.	Efforts	to	disentangle	
these	networks	are	sometimes	complicated	by	the	limited	ability	to	culture	partner	species	
independently	and	underdeveloped	molecular	toolkits.	As	an	example,	cyanobionts	are	
vertically	transmitted	in	the	 Nostoc-Azolla	mutualism.	This	means	that	the	association	
between	partners	is	extremely	long-lived	and	stable	across	generations,	to	the	extent	that	it	
is	debatable	if	de	novo	infection	occurs	in	a	natural	context,	and	if	free-living	cyanobacteria	
can	grow	independently	(Adams	et	al.	2013).	While	this	level	of	stability	would	be	desirable	
	

13	

	

to	replicate	in	engineered	microbial	interactions,	it	renders	natural	mutualisms	with	this	
level	of	interdependency	difficult	to	dissect.		

Part	II:	Promise	and	current	limitations	of	the	application	of	synthetic	microbial	

communities	

Synthetic	microbial	ecology	&	microbial	ecology	theory	

Synthetic	microbial	ecology	offers	an	alternative	approach	to	study	fundamental	
questions	concerning	microbial	interactions	and	to	examine	how	local	interactions	
between	microbes	can	lead	to	complex	higher-order	patterns	at	the	population	level.	
Synthetic	microbial	ecology	uses	simple	artificial	communities	that	retain	features	of	
natural	microbial	communities,	but	which	display	greatly	reduced	network	complexity	in	
terms	of	the	number	of	interacting	species	and	the	degree	of	connectivity	between	species	
(Momeni	et	al.	2011).	Synthetic	microbial	communities	are	typically	established	between	
experimentally-tractable	organisms	that	can	be	selected	or	engineered	to	interact	through	
defined	pathways.	The	ability	to	construct	ecologies	composed	of	model	organisms	can	be	
beneficial	for	a	variety	of	reasons	including	the	short	generation	times,	small	genomes,	
advanced	genetic	toolkits,	and	capacity	to	freeze	populations	for	evolution	studies.	There	
are,	of	course,	inherent	tradeoffs	between	the	level	of	control	over	such	artificial	systems	
and	the	realism	of	these	platforms	to	natural	microbial	communities	(Figure	1.1).		

	

14	

	

15	

	

Figure	1.1:	Study	of	microbial	ecology	at	different	levels	of	abstraction		

Understanding	the	factors	that	shape	population-level	behaviors	of	microbial	communities	
requires	independent	lines	of	approach.	Natural	microbial	communities	are	frequently	
composed	of	dozens	to	hundreds	of	different	microbial	species,	many	of	which	may	have	
co-evolved	in	the	given	ecosystem	over	many	generations.	Direct	(solid	lines)	and	indirect	
(dotted	lines)	interactions	comprise	a	complex	network	between	species.	Natural	
environments	also	display	highly	irregular	physical	and	chemical	properties,	which	can	
dynamically	shift	over	time.	Different	degrees	of	reductionism	have	been	used	to	
disentangle	the	complexities	of	natural	systems.	Mathematical	theories	and	computational	
models	represent	the	most	abstracted	field	of	research,	including	low-resolution	
population-based	models,	metabolic	network	models,	and	individual-based	simulations	
(Song	et	al.	2014a).	Pure	microbial	cultures	allow	detailed	physiological	studies	of	an	
isolated	species	in	a	highly	defined	and	homogeneous	environment.	Depending	on	the	
microbe,	a	variety	of	genetic	tools	and	sophisticated	circuits	may	assist	analysis.	Synthetic	
microbial	consortia	consist	of	2	or	more	microbes	that	interact	through	defined	pathways.	
Although	other	emergent	interactions	(red	dotted	lines)	are	likely	to	arise	between	species,	
the	networks	are	regarded	to	be	much	less	tightly	integrated,	and	the	nature	of	such	
interactions	can	be	probed	through	genetic	approaches	generally	not	available	during	the	
study	of	natural	ecologies.	
		Yet,	it	is	recognized	that	the	field	of	microbial	ecology	currently	has	limited	theories	that	
can	be	used	to	predict	the	behavior	of	populations	(Prosser	et	al.	2007;	Widder	et	al.	2016),	
	

	

and	artificial	communities	are	increasingly	regarded	as	an	important	bridge	between	
abstract	mathematical	models	and	the	complexity	of	natural	consortia.	For	some	of	these	
reasons,	the	last	decades	have	seen	a	steady	rise	in	the	use	of	synthetic	microbial	
communities	as	a	method	to	study	fundamental	questions	related	to	the	structure	and	
function	of	ecological	networks	(Jessup	et	al.	2005).		
In	recent	years,	study	of	synthetic	microbial	ecologies	has	provided	a	number	of	
useful	platforms	for	study	of	a	variety	of	variables	known	to	influence	natural	microbial	
consortia	(De	Roy	et	al.,	2014;	Jessup	et	al.,	2004;	Prosser	et	al.,	2007;	Figure	1.1).	Natural	
consortia	can	be	composed	of	hundreds	to	thousands	of	species	that	interact	through	the	
exchange	of	(often	unknown)	metabolites	and	signaling	molecules,	yet	despite	these	
intricate	features,	such	communities	display	a	surprising	persistence	and	resilience	in	the	
face	of	environmental	perturbations.	This	trait	is	commonly	referred	to	as	robustness,	or	
the	ability	of	a	community	to	maintain	its	functional	and	structural	integrity	in	the	face	of	
fluctuating	environmental	and	biotic	conditions.	In	previously-discussed	examples,	the	
Azolla-Nostoc	mutualism	can	persist	through	numerous	generations	(Adams	et	al.	2013),	
while	 Prochlorococcus-dominated	communities	in	the	open	ocean	display	a	surprising	
degree	of	regularity	in	composition	from	year	to	year,	even	cycling	through	predictable,	
seasonally-driven	states	that	maintain	key	features	at	the	population	level	(Malmstrom	et	
al.	2010).	One	core	ecological	theory	is	that	higher	diversity	contributes	to	increased	
community	stability;	robustness	is	derived	in	part	by	diversity	and	by	redundancy	of	
functions	divided	amongst	multiple	community	members	(Ives	and	Carpenter,	2007;	
McCann,	2000).	These	theories	are	supported	by	studies	that	used	a	recombinatorial	
approach	to	assemble	microbial	communities	with	a	variable	number	of	phototrophs,	
	

16	

	
heterotrophic	bacteria,	and	predators,	finding	that	community	level	features	(such	as	total	
CO2	flux	or	biomass	production)	became	more	reliable	with	increased	species	diversity	
(McGrady-Steed	et	al.	1997;	Naeem	and	Li	1997).		
	
Other	studies	have	emphasized	the	role	of	spatial	structuring	in	promoting	
community	stability.	For	example,	multiple	synthetic	microbial	systems	for	studying	
predator-prey	interactions	have	been	developed	and	examined	for	conditions	that	
influence	the	rate	of	extinction	of	one	or	more	partners	(Bohannan	and	Lenski	2000;	Kerr	
et	al.	2002).	In	one	notable	example,	artificial	communities	of	protists	were	examined	for	
stability	under	homogenous	environments,	or	under	conditions	where	the	same	total	
population	size	could	become	subdivided	into	connected,	but	locally-differentiated	
microcosms	(Holyoak	and	Lawler	1996).	The	results	indicated	that	the	heterogenous	
environment	could	substantially	stabilize	the	artificial	microbial	community,	providing	
evidence	in	support	of	metapopulation	theories	that	had	primarily	been	examined	in	
mathematical	models	(Hanski	and	Hanski	1998).	Similarly,	a	spatially	structured	
environment	
cooperative	 behaviors	
in	
microenvironments	that	contain	non-cooperating	individuals	(Doebeli	and	Hauert	2005).	
In	a	homogenous	environment,	cooperative	behaviors	that	benefit	neighboring	species	but	
which	incur	a	fitness	cost	upon	the	individual	(e.g.,	secretion	of	a	metabolite	that	requires	
investment	of	biochemical	resources)	are	often	counter-selected,	since	non-cooperating	
community	members	can	reap	public	benefits	without	the	costs	of	contributing.	Both	
theoretical	 and	 experimental	 evidence	 using	 synthetic	 microbial	 consortia	 have	
demonstrated	that	structuring	microbial	partners	into	localized	communities	can	stabilize	
cooperative	behavior	(Kim	et	al.	2008;	Chuang	et	al.	2009;	Waite	and	Shou	2012;	Allen	et	al.	
	

resilience	 of	

can	 also	

increase	

the	

17	

	

2013;	Momeni	et	al.	2013;	Kelsic	et	al.	2015;	Pande	et	al.	2016).	Briefly,	when	exchange	of	
goods	between	partners	is	partially	restricted	to	localized	environments	(e.g.,	within	flocs,	
or	isolated	colonies),	isolated	populations	dominated	by	non-cooperative	individuals	have	
weak	positive	feedback	loops,	while	nearby	micro-communities	dominated	by	cooperators	
can	have	robust	positive	feedback	and	relatively	high	total	growth	rates.		
	
The	above	examples	serve	to	demonstrate	how	the	simplified	composition	and	
increased	molecular	tools	of	synthetic	microbial	consortia	enable	testing	of	fundamental	
theories	of	microbial	ecology.	Interested	readers	are	directed	to	several	excellent	reviews	
for	additional	information	on	synthetic	microbial	ecology	(Jessup	et	al.	2004;	Prosser	et	al.	
2007;	Kazamia	et	al.	2012a;	De	Roy	et	al.	2014;	Widder	et	al.	2016).		

The	biotechnological	potential	of	synthetic	consortia		

literature	

As	 with	 the	 academic	

in	 microbiology,	 the	 majority	 of	 current	
bioindustrial	technologies	rely	on	microbial	monocultures,	although	polyculture	offers	
several	potential	benefits.	In	particular,	if	the	high	metabolic	efficiencies	and	robustness	
commonly	observed	in	natural	consortia	could	be	replicated	in	synthetic	microbial	
communities,	it	would	have	considerable	implications	for	a	wide	array	of	industrial,	
medical,	and	environmental	applications	(Goers	et	al.	2014).	Although	one	common	
conception	in	synthetic	biology	is	that	improved	genetic	tools	will	allow	us	to	reprogram	a	
target	biological	 chassis	(e.g.,	 E.	 coli)	for	any	desired	output,	many	ecological	examples	
argue	that	mixed	communities	should	typically	outperform	a	single	species	(no	matter	how	
extensively	engineered)	in	terms	of	total	bioproductivty	and	robustness.		
	

18	

and	

Savage	 2018),	

(Chaijarasphong	

compartmentalization	

The	concept	of	biological	division	of	labor	is	chief	among	the	reasons	that	increased	
bioproductivity	can	be	observed	in	consortia	(Brenner	et	al.	2008;	Werner	et	al.	2014;	Hays	
et	al.	2015;	Lindemann	et	al.	2016).	Individual	members	in	a	complex	community	adopt	
specialized	roles,	allowing	niche	differentiation	and	functional	complementarity	that	can	
enable	more	efficient	resource	utilization	(Savage	et	al.	2007).	This	is	reflected	in	higher	
bioproductivity	yields	(e.g.,	total	biomass)	from	communities	compared	to	populations	
containing	one	species,	an	observation	that	also	applies	in	rationally	designed	systems	
(Eiteman	et	al.	2008;	Shong	et	al.	2012).	Indeed,	the	improved	efficiency	of	networks	
containing	specialists	is	a	core	tenant	of	biological	systems	at	other	scales,	including	
intracellular	
tissue	
differentiation	in	multicellular	organisms	(Ispolatov	et	al.	2012),	and	even	within	social	
economic	theories	(Werner	et	al.	2014).	Furthermore,	compartmentalization	of	metabolic	
reactions	across	distinct	species	can	mitigate	constraints	imposed	by	tradeoffs	between	
different	objectives,	where	increased	efficiency	in	one	objective	(e.g.,	metabolic	reaction),	
comes	at	the	cost	of	another.	For	example,	different	enzymes	may	compete	for	a	common	
pool	of	biomolecules	(e.g.,	ATP,	NADH,	or	other	precursors	and	co-factors)	and	inter-
enzyme	competition	(even	between	enzymes	within	a	simple,	linear	metabolic	pathway),	
can	result	in	the	accumulation	of	intermediates	and/or	constraints	in	the	total	flux	through	
desired	metabolic	steps	(Lindemann	et	al.	2016).	In	other	instances,	incompatibilities	
between	one	metabolic	pathway	and	another	can	essentially	preclude	their	simultaneous	
operation.	One	example	of	this	is	represented	by	attempts	to	produce	hydrogen	gas	by	
using	reductant	from	oxygenic	photosynthesis:	hydrogenase	enzymes	are	highly	oxygen-
sensitive,	generating	many	complications	when	both	processes	are	confined	within	the	
	

	

19	

	

same	cell	(Ghirardi	2015;	Posewitz	2018).	Similarly,	incompatibilities	in	metabolic	
regulation	have	made	it	difficult	to	engineer	a	single	heterotrophic	microbe	that	can	
efficiently	utilize	all	major	sugars	released	from	the	hydrolysis	of	lignocellulosic	materials	
(Eiteman	et	al.	2008;	Minty	et	al.	2013).		
Robustness	of	natural	systems	is	another	feature	that	is	particularly	important	for	
biotechnological	 applications	 where	 environmental	 conditions	 cannot	 be	 strictly	
controlled.	One	application	where	this	is	the	case	is	in	the	scaled	cultivation	of	algae	and	
cyanobacteria	for	the	production	of	fuels,	polymers,	and	other	biologics.	Algal	farming	
requires	increased	scaling	of	the	surface	area	to	expand	the	available	input	light,	which	
complicates	the	design	of	economically-feasible	enclosed	photobioreactors,	especially	for	
the	production	of	commodity	goods	(Ducat	et	al.	2011;	Chisti	2013).	By	contrast,	the	open	
systems	that	are	deemed	more	realistic	(Sheehan	et	al.	1998;	Chisti	2007)	make	algal	
cultures	vulnerable	to	invasive	species	and	largely	preclude	the	possibility	of	tight	
environmental	control	(e.g.,	temperature;	Gupta	et	al.,	2015).	Any	other	biotechnological	
applications	that	require	release	into	natural	environments	(e.g.,	the	application	of	soil	
microbes	to	improve	crop	yields;	Chaparro	et	al.,	2012),	will	also	face	the	challenge	of	
obtaining	a	predictable	output	under	highly	variable	biotic	and	abiotic	conditions.		
Most	microbial	species	are	not	sufficiently	robust	to	be	used	in	a	monoculture	
approach	for	applications	that	are	exposed	to	natural	environmental	fluctuations.	Among	
the	largest	problems	of	scaled	cultivation	of	algae	and	cyanobacteria	is	the	high	rate	of	
“pond	crashes”	caused	by	invasion	of	a	foreign	microbe	or	virus	(Smith	et	al.	2010;	Wang	et	
al.	2013;	Carney	and	Lane	2014).	By	contrast,	natural	ecosystems	with	diverse	members	
can	be	described	as	reaching	a	“climax”	steady	state,	where	multiple	stable	equilibria	can	
	

20	

Synthetic	co-cultures	for	photosynthesis-driven	bioindustry	

be	reached	following	a	perturbation,	helping	to	minimize	variation	and	invasion	by	foreign	
species	(May	1977;	Law	and	Daniel	Morton	1996).	If	effective	ecological	principles	can	be	
identified	and	applied	to	synthetic	consortia,	higher	productivities	and	“self-regulating”	
behaviors	should	be	theoretically	possible	to	engineer	for	biotechnological	applications.	
Given	the	improbability	of	maintaining	axenic	ponds	of	cyanobacteria	or	algae,	
	
increasing	focus	is	being	placed	on	identifying	suitable	partner	species	that	will	improve	
productivity	or	stability	of	the	culture.	Many	of	the	current	efforts	involve	prospecting	for	
heterotrophic	bacteria	which	increase	total	culture	yields.	As	an	example,	co-inoculation	of	
a	bacterial	species	(Brevundimonas	sp.)	that	was	isolated	as	a	contaminant	of	algal	cultures,	
with	 Chlorella	 ellipsoidea	resulted	in	up	to	3	times	greater	algal	growth	than	that	of	 C.	
ellipsoidea	alone	(Park	et	al.	2008).	There	are	many	similar	examples	(Amin	et	al.	2009;	
Natrah	et	al.	2014;	Cho	et	al.	2015),	and	microbial	species	that	promote	growth	of	
cyanobacteria	and	algae	need	not	be	isolated	from	environments	where	they	grow	in	
proximity	to	these	phototrophs	(Hays	et	al.	2017;	Tandon	and	Jin	2017).	The	fact	that	
species	that	do	not	naturally	co-exist	with	a	given	phototrophic	microbe	can	also	enhance	
growth	in	co-culture	suggests	that	there	may	be	generalizable	pathways	by	which	
heterotrophs	perform	beneficial	functions	for	algae	and	cyanobacteria.	These	pathways	
may	include:	mitigation	of	ROS,	increasing	dissolved	CO2,	cross-feeding	of	TCA	cycle	
intermediates,	increasing	accessibility	of	essential	metal	ions	(i.e.,	via	siderophore	
secretion),	or	secretion	of	soluble	vitamins	(Natrah	et	al.	2014;	Cooper	and	Smith	2015;	
Hays	et	al.	2017).	Identifying	a	more	complete	list	of	common	synergistic	interactions	
	

	

21	

	

between	autotroph/heterotroph	pairs	could	benefit	from	the	use	of	synthetic	communities	
where	each	partner	species	has	a	developed	genetic	toolkit	and	robust	metabolic	models.	
Deeper	understanding	of	naturally-occurring	interactions	at	the	species	level	could	assist	
in	the	rational	design	of	partner	microbes	that	could	optimize	an	artificial	system	(e.g.,	
production	in	a	race-way	pond)	towards	target	goals.	
One	systematic	way	to	explore	the	design	space	of	synthetic	microbial	communities	
is	through	the	development	of	modular	consortia	(Figure	1.2).	

Figure	1.2:	Modularity	in	microbial	consortia		

	
Synthetic	microbial	consortia	can	be	designed	in	a	modular	fashion,	where	each	“module”	
contributes	one	or	more	key	functions	towards	the	overall	capabilities	of	the	community.	
Compatible	modules	may	be	recombined	with	one	another	in	a	flexible,	‘plug	&	play’	
approach	to	generate	a	range	of	related	communities	with	distinct	outputs.	This	approach	
has	the	advantage	of	allowing	cross-comparison	of	similar	synthetic	consortia	to	identify	
common	features	and	interaction	patterns	shared	across	different	species	pairings.	By	
contrast,	natural	microbial	communities	are	composed	of	hundreds	to	thousands	of	
species:	though	individual	partners	may	be	arbitrarily	categorized	(e.g.,	by	trophic	level),		
	

22	

	

Figure	1.2	(cont’d)	
there	is	a	high	degree	of	functional	redundancy.	Furthermore,	because	members	of	these	
communities	have	co-evolved	over	many	generations,	interspecies	interactions	are	more	
likely	to	be	context-specific.	This	can	complicate	the	ability	to	define	discrete	roles	or	
identify	predictable	interaction	patterns	for	individual	species.	As	the	field	of	synthetic	
microbial	ecology	matures,	construction	of	modular	communities	that	more	closely	mimic	
natural	systems	(e.g.,	composed	of	species	with	specialized	metabolic	modes)	may	assist	in	
the	identification	of	generalizable	design	principles.	
	
	In	synthetic	biology,	a	biological	module	is	defined	as	a	unit	of	function	that	can	be	
separated	from	its	native	context	and	repurposed	in	new	networks	while	retaining	fidelity	
of	core	functions	(Andrianantoandro	et	al.	2006).	While	it	is	a	common	practice	of	synthetic	
biology	to	conceptualize	biological	units	as	‘modular’	on	the	molecular	scale	(e.g.,	protein	
domains,	genes,	promoter	elements),	these	concepts	are	increasingly	being	applied	at	the	
level	of	whole	cells,	tissues	in	multicellular	organisms,	or	species	within	ecosystems	(Ortiz-
Marquez	et	al.	2013;	Cameron	et	al.	2014;	Ducat	2017).	In	this	context,	a	given	strain	within	
a	microbial	community	would	serve	a	defined	set	of	functions,	but	could	be	substituted	in	a	
“plug-and-play”	fashion	with	a	different	species	that	fulfills	those	roles.	One	advantage	of	
developing	modular	microbial	communities	is	that	it	streamlines	design	of	new	multi-
species	consortia,	and	facilitates	the	identification	of	common	themes	of	interaction	
between	related	communities.	
Some	 recent	 examples	 of	 modular	 microbial	 platforms	 are	 based	 off	 of	
cyanobacteria	that	have	been	engineered	to	secrete	soluble	carbohydrates.	These	
engineered	strains	can	behave	as	a	‘carbon	fixation	module’	(Figure	1.2)	within	synthetic	
communities	that	is	analogous	to	natural	cyanobacterial	symbionts	that	secrete	organic	
carbon	for	the	community,	as	Prochlorococcus	does	in	marine	ecosystems	(see	above).	One	
example	that	has	been	utilized	in	numerous	synthetic	co-cultures	is	a	strain	of	
	

23	

Synechococcus	 elongatus	 sp.	 PCC	 7942	 that	 has	 been	 modified	 to	 express	 the	
sucrose/proton	symporter,	 cscB	(Ducat	et	al.	2012).	This	strain	stably	exports	a	large	
proportion	of	photosynthetically-fixed	carbon	as	the	easily	metabolized	disaccharide,	
sucrose	(Ducat	et	al.	2012;	Abramson	et	al.	2016).	Multiple	labs	have	shown	that	
photosynthate	from	these	cyanobacteria	is	sufficient	as	the	sole	source	of	carbon	for	
metabolism	and	growth	of	co-cultured	heterotrophic	microbes.	In	this	design,	the	
heterotroph	can	be	conceptualized	as	a	‘conversion	module’	to	transform	the	fixed	carbon	
into	more	valuable	bioproducts,	including	the	bioplastic	precursors	polyhydroxyalkanoate	
(paired	microbe:	Psuedomonas	putida;	Löwe	et	al.,	2017),	polyhydroxybuterate	(Halomonas	
bolieviensis;	Weiss	and	Ducat,	2017,	 E.	 coli;	Hays	et	al.,	2017,	or	 Azotobacter	 vinelandii;	
Smith	and	Francis,	2016),	fatty	acids	(S.	cerevisiae	or	Rhodotorula	glutinis;	Li	et	al.,	2017),	
or	secreted	enzymes	(Bacillus	subtilis;	Hays	et	al.,	2017).	Of	note,	in	a	recent	report	utilizing	
synthetic	co-cultures	of	sucrose-secreting	 S.	 elongatus	and	 H.	 bolieviensis,	the	designed	
community	was	able	to	continuously	produce	PHB	over	the	course	of	more	than	5	months	
while	also	resisting	invasion	by	a	common	laboratory	contaminant	(Weiss	and	Ducat	
2017).	Collectively,	these	results	highlight	the	flexibility	and	utility	of	adopting	a	modular	
approach	to	synthetic	consortia	design	(Figure	1.2).	
A	number	of	other	cyanobacterial	strains	have	been	engineered	to	export	
the	
in	
the	
carbohydrates,	 permitting	 substitution	 of	
aforementioned	synthetic	consortia.	For	example,	 cscB	has	been	expressed	to	improve	
carbohydrate	secretion	in	numerous	other	model	cyanobacteria	(Du	et	al.	2013;	Duan	et	al.	
2016;	Song	et	al.	2016).	Alternatively,	cyanobacteria	have	been	engineered	to	secrete	a	
variety	of	other	carbohydrates	(Niederholtmeyer	et	al.	2010;	Xu	et	al.	2012;	Aikens	and	
	

“photosynthetic	 module”	

	

24	

	

Turner	2013;	McEwen	et	al.	2013;	Hays	and	Ducat	2015),	whereas	some	microalgal	strains	
have	been	engineered	to	secrete	glycerol	(Demmig-Adams	et	al.	2014).	Additionally,	many	
cyanobacterial	mutants	that	are	deficient	in	storing	carbohydrates	through	glycogen	
synthesis	instead	secrete	a	wide	array	of	carbon	compounds	as	overflow	metabolism	
products	(Carrieri	et	al.	2012;	Gründel	et	al.	2012;	Xu	et	al.	2012;	Hickman	et	al.	2013).			
Other	modular	synthetic	microbial	communities	have	been	generated	using	strains	
engineered	to	fix	atmospheric	nitrogen	and	provide	it	to	neighboring	species.	Some	such	
experiments	were	originally	performed	with	the	diazatrophic	cyanobacteria,	 Anabaena	
variabilis,	specifically	utilizing	mutants	where	the	nitrogenase	genes	are	constitutively	
derepressed	(Spiller	et	al.	1986).	Consequently,	these	strains	maintain	nitrogenase	activity	
even	after	fulfilling	their	own	needs,	resulting	in	the	secretion	of	excess	ammonia	(Singh	
and	Tiwari	1998).	When	such	mutants	were	co-cultivated	with	wheat	or	rice,	they	
supplemented	the	nitrogen	requirements	of	the	plant	and	increased	crop	yields	(Latorre	et	
al.	1986;	Spiller	and	Gunasekaran	1990).	More	recently,	a	heterotrophic	diazotrophic	
species,	 A.	 vinelandii,	was	modified	to	continuously	express	nitrogenase	and	secrete	
ammonia,	and	these	strains	have	been	used	in	co-culture	with	cyanobacteria,	algae	and	
with	plants	(Ortiz-Marquez	et	al.	2012,	2014;	Smith	and	Francis	2016;	Ambrosio	et	al.	
2017).	In	each	case,	the	partner	species	effectively	gained	the	benefit	of	nitrogen-fixing	
leading	 to	 enhanced	 co-culture	 productivity.	
capabilities	 through	 the	 association,	
Experiments	in	development	seek	to	combine	both	the	carbon-	and	nitrogen-	fixing	
modules	into	a	single	species	by	heterologously	expressing	cscB	within	Anabaena	strains	
that	also	secrete	ammonia:	such	cyanobacterial	strains	are	regarded	as	promising	for	
	

25	

	

supporting	microbial	transformations	during	long-range	space	flights	(Verseux	et	al.	2015,	
2016).			
A	final	class	of	rationally-designed	interactions	between	a	phototroph	and	a	
	
heterotroph	involves	the	exchange	of	vitamins	or	other	cofactors.	Again,	co-dependencies	
can	be	programmed	into	selected	partners	that	mimic	exchanges	that	are	routinely	found	
in	natural	environments.	In	co-cultures	of	the	green	algae	 Lobomonas	 rostrata	with	the	
rhizobial	bacterium	Mesorhizobium	loti,	algae	could	secrete	sufficient	carbon	to	support	the	
growth	of	the	prokaryote,	while	receiving	sufficient	cobalamin	(vitamin	B12)	for	its	own	
needs	 (Kazamia	 et	 al.	 2012b).	
the	 mixed	
autotroph/heterotroph	populations	exhibited	improved	bioproduction	and/or	metabolic	
functionalities	were	not	available	to	or	achievable	by	the	individual	species	alone.	
Furthermore,	many	of	these	platforms	function	as	simplified	systems	to	study	key	
exchanges	that	are	found	in	natural	communities	in	a	more	methodical	fashion.	

the	 above	 consortia,	

In	 each	 of	

Limitations	in	synthetic	co-culture	approaches	and	future	perspectives	

Despite	a	number	of	recent	advances,	the	use	of	synthetic	consortia	for	both	
academic	and	applied	research	is	still	in	its	infancy	and	must	overcome	a	number	of	
limitations	to	realize	their	potential.	Ironically,	while	natural	consortia	display	a	high	
degree	of	robustness,	the	simplified	microbial	communities	built	to	mirror	them	are	often	
highly	unstable	even	under	controlled	conditions.	Furthermore,	the	majority	of	published	
examples	of	synthetic	consortia	are	composed	of	only	two	heterotrophic	strains,	often	that	
are	same	species	but	which	have	been	slightly	modified	with	different	genetic	constructs	
(e.g.;	Basu	et	al.,	2005;	Danino	et	al.,	2010;	Goers	et	al.,	2014;	Wintermute	and	Silver,	2010).	
	

26	

(i.e.,	

chemoheterotrophs,	 methylotrophs,	

Natural	communities	are	composed	of	individuals	from	many	species	and	with	distinct	
trophic	 modes	
photoautotrophs,	
photomixotrophs,	and	chemoautotrophs).	The	presence	of	metabolic	specialists	can	be	
anticipated	to	have	profound	effects	on	community	stability	in	the	face	of	environmental	
perturbations.	Indeed,	a	cornerstone	ecological	principle	is	that	species	co-existence	in	the	
long-term	is	impossible	if	they	share	the	same	resources	and	niches	(MacArthur	and	Levins	
1964;	Kazamia	et	al.	2012a).	Thus,	increasing	the	number	of	synthetic	microbial	platforms	
that	are	composed	of	more	than	one	trophic	mode	may	partially	alleviate	the	problem	of	
instability.	Providing	a	structured	environment	is	another	approach	that	may	be	used	to	
increase	the	robustness	of	synthetic	consortia.	As	discussed	above,	there	are	a	number	of	
theoretical	and	experimental	studies	that	demonstrate	that	the	physical	structure	and	
spatial	arrangement	between	microbial	partners	can	increase	the	long-term	stability	of	
interactions	that	are	otherwise	prone	to	collapse	(Kim	et	al.	2008;	Chuang	et	al.	2009;	
Waite	and	Shou	2012;	Allen	et	al.	2013;	Momeni	et	al.	2013;	Kelsic	et	al.	2015;	Pande	et	al.	
2016).	For	example,	a	tripartite	heterotrophic	community	that	was	unstable	when	
cultivated	in	well-mixed	homogenous	environments	could	be	stabilized	over	long	time	
periods	by	sequestration	of	each	species	into	distinct	wells	of	a	microfluidic	device	that	
were	connected	by	channels	allowing	exchange	of	small	molecules	(Kim	et	al.	2008).	
Microfluidic	devices	also	may	greatly	increase	the	capacity	to	test	many	microbial	consortia	
through	recombinatorial	approaches	by	miniaturizing	growth	chambers	(Nai	and	Meyer	
2018).	Yet,	for	some	large-scale	applications	(such	as	the	microalgal	ponds	discussed	
above)	cultivation	in	such	carefully	manufactured	conditions	may	not	be	economically	
	

	

27	

	

realistic.	An	alternative	approach	to	providing	structure	within	microbial	communities	
would	be	to	engineer	the	cells	to	self-organize	into	higher	order	patterns.	One	avenue	
worth	additional	exploration	in	this	regard	is	the	refinement	of	cell-cell	attachment	
systems	(Fedeson	and	Ducat	2016;	Zhang	et	al.	2017;	Glass	and	Riedel-Kruse	2018).	
Programmable	cell	adhesion	between	partner	species	could	allow	biologists	to	better	
define	the	spatial	positioning	and	architecture	of	the	community	(e.g.,	programming	cells	to	
flocculate,	or	form	structured	biofilms),	gaining	the	benefits	of	a	structured	environment	
even	in	an	otherwise	homogenized	environment.	
Finally,	the	instability	of	many	synthetic	microbial	communities	limits	the	time	
scales	these	consortia	have	been	observed	(usually	hours	to	a	couple	of	days;	Goers	et	al.,	
2014).	Increasing	the	period	of	observation	for	stable	synthetic	communities	would	offer	a	
new	window	into	the	early	evolution	of	symbiotic	interactions.	While	we	have	learned	a	
great	deal	from	“top	down”	dissection	of	specific	natural	mutualisms	(e.g.,	the	chloroplast),	
the	prehistoric	origins	of	these	relationships	can	only	be	inferred.	By	contrast,	because	the	
partner	species	in	synthetic	microbial	communities	are	naïve	with	regard	to	one	another,	
stable	synthetic	ecologies	offer	a	“bottom-up”	approach	to	study	the	early	stages	of	
evolution	of	a	symbiosis	(Hom	and	Murray	2014).	For	example,	Shou	et.	al.	have	studied	the	
evolution	of	synthetic	yeast	consortia	in	the	early	generations	to	determine	what	adaptive	
traits	underlie	the	observed	capacity	of	these	populations	to	evolve	towards	higher	
population	density	limits	(Shou	et	al.	2007).	Long-term	study	of	the	co-evolution	of	species	
within	synthetic	autotroph/heterotroph	consortia	could	provide	new	insights	into	common	
themes	of	the	early	stages	of	the	formation	of	cyanobacterial	and	algal	mutualisms	that	are	
abundant	in	nature.	
	

28	

	

Concluding	remarks	

	
Synthetic	microbial	consortia	are	increasingly	being	used	to	disentangle	the	
complex	networks	of	natural	microbial	communities	and	program	consortia	to	efficiently	
perform	valuable	services.	Synthetic	microbial	consortia	offer	a	biological	platform	for	
probing	the	mechanisms	of	microbial	interactions	that	is	simpler	and	more	easily	
controlled	than	natural	microbial	ecologies.	The	earliest	examples	of	synthetic	microbial	
consortia	have	already	provided	a	wealth	of	information	useful	for	the	development	of	
generalizable	theories	of	microbial	ecology.	Yet,	these	early	examples	are	often	overly	
simplistic,	with	limited	numbers	of	metabolic	specialists	and	poor	stability.	Improvement	
upon	early	examples	may	help	to	determining	effective	design	rules	that	increase	the	
robustness	and	bioproductivity	of	engineered	consortia.	The	ability	to	confer	these	traits	
on	engineered	consortia	would	greatly	expand	the	viability	of	many	microbiology	
applications	 across	 diverse	
therapies,	
living	
bioremediation,	and	production	of	sustainable	fuels	and	other	commodity	products.		

including	 human	 medicine,	

fields,	

	

	

	

29	

CHAPTER	2.1:	CYANOBACTERIAL	SURFACE	DISPLAY	SYSTEM	MEDIATES	ENGINEERED	

INTERSPECIES	AND	ABIOTIC	BINDING	

The	work	presented	in	this	chapter	has	been	published:		

	
			
		

30	

Reprinted	with	permission	from	(Fedeson,	D.	T.	&	Ducat,	D.	C.	Cyanobacterial	surface	
display	system	mediates	engineered	interspecies	and	abiotic	binding.	ACS	Synth.	Biol.	6,	
367–374	(2016).).	Copyright	©	2016	American	Chemical	Society.	
	

	

	

	

	

	

Abstract	

Cyanobacteria	are	uniquely	suited	for	development	of	sustainable	bioproduction	
platforms,	but	are	currently	underutilized	in	scaled	applications	in	part	due	to	a	lack	of	
genetic	tools.	Here,	we	develop	a	surface	display	system	in	the	cyanobacterial	model	
Synechococcus	 elongatus	PCC7942	via	expression	of	modified	versions	of	the	outer	
membrane	porin,	SomA.	Importantly,	we	demonstrate	accessibility	of	heterologous	
functional	groups	on	the	recombinant	porin	to	the	external	environment	in	living	cells.	We	
show	that	this	requires	the	removal	of	occluding	factors	that	include	lipopolysaccharides	
and	a	putative	surface	layer	protein.	Displayed	epitopes	on	SomA	can	be	utilized	to	mediate	
physical	adhesion	between	living	cyanobacteria	and	abiotic	surfaces	or	an	engineered	
Saccharomyces	cerevisiae	partner	strain.	We	show	that	>80%	of	cyanobacterial	cells	attach	
to	functionalized	magnetic	beads,	allowing	for	magnet-assisted	recovery.	This	work	
showcases	the	development	of	a	functional	surface	display	system	in	cyanobacteria,	with	
wide-ranging	applications.	
	

	

	

31	

	

Introduction	

Cyanobacteria	and	algae	are	increasingly	examined	as	platforms	for	solar-driven	
bioproduction	due	to	their	favorable	traits	relative	to	traditional	plant	crops.	Current	
photobiological	production	schemes	revolve	around	microbial	conversion	of	plant-derived	
sugars	(Wyman	and	Goodman	1993;	Papini	et	al.	2010),	creating	sustainability	concerns	
regarding	arable	land	and	potable	water	use	(Dismukes	et	al.	2008).	Cyanobacteria	are	an	
alternative	crop	species	that	could	disentangle	bioindustrial	production	from	agricultural	
processes	(Ducat	et	al.	2011).	Furthermore,	cyanobacteria	have	advantages	over	other	crop	
species	including	their	high	photosynthetic	efficiency,	genetic	tractability,	and	minimal	
nutritional	requirements	(Castenholz	1988;	Dismukes	et	al.	2008;	Rittmann	2008;	Rastogi	
and	Sinha	2009;	Ducat	et	al.	2011;	Kiran	et	al.	2014).	Despite	considerable	promise,	
technical	and	economic	barriers	to	cultivating	cyanobacteria	have	limited	their	scaled	
applications	(Ortiz-Marquez	et	al.	2013).	
Harvesting	cellular	biomass	is	a	key	concern	in	the	economics	of	scaled	
cyanobacterial	cultivation.	The	size,	charge,	and	density	of	cyanobacteria	allow	them	to	
remain	suspended	in	stagnant	media,	and	therefore	biomass	must	be	harvested	through	
flocculation-induced	sedimentation,	filtration,	or	centrifugation	(Grima	et	al.	2003).	Initial	
biomass	recovery	and	subsequent	dewatering	techniques	are	energy	intensive,	limit	
harvesting	rates,	and	can	account	for	~20-30%	of	total	production	costs	(Grima	et	al.	
2003).	Rather	than	using	expensive	infrastructural	approaches,	these	costs	could	be	
potentially	mitigated	or	eliminated	through	use	of	engineered	cyanobacteria	strains.	
However,	the	relatively	limited	cyanobacterial	molecular	toolkit	inhibits	our	ability	to	
provide	biological	solutions	(Ortiz-Marquez	et	al.	2013).		
	

32	

	

Surface	display	is	a	molecular	tool	relevant	to	scaled	biomass	recovery	that	has	
broad	applications	in	academic	and	industrial	pursuits.	For	example,	surface	display	
systems	have	been	utilized	for	numerous	applications	in	bacteria	and	eukaryotic	
microorganisms,	including:	bioremediation,	live	vaccines,	biocatalysts,	and	directed	
evolution	(Samuelson	et	al.	2002;	Kondo	and	Ueda	2004).	A	functionalizable	cyanobacterial	
surface	would	permit	the	display	of	useful	enzymatic/chemical	groups,	allowing	for	
selective	biomass	harvesting	techniques	and	other	applications.	Previous	attempts	have	
been	made	to	functionalize	the	surface	of	cyanobacterial	cells,	but	were	met	with	only	
partial	success	(Chungjatupornchai	and	Fa-Aroonsawat	2008;	Chungjatupornchai	and	Fa-
aroonsawat	2009;	Chungjatupornchai	et	al.	2011;	Ferri	et	al.	2015),	unable	to	demonstrate	
that	the	sequences	targeted	for	surface	display	were	operative,	appropriately	localized,	and	
accessible.	An	effective	surface	display	system	for	any	application	requires	the	target	
protein	to	localize	to	the	outermost	surface	of	the	cell	and	for	this	functional	moiety	to	be	
physically	accessible	to	potential	interaction	partners	and	substrates.		
Here,	we	conclusively	demonstrate	the	targeting	of	a	heterologous	peptide	to	the	
surface	of	 Synechococcus	 elongatus	PCC	7942.	We	systematically	investigate	potential	
limitations	to	effective	surface	display	(e.g.,	protein	 expression,	
localization,	 and	
accessibility)	to	determine	the	major	limiting	factors	that	have	likely	prevented	successful	
implementation	of	surface	display	in	previous	efforts.	We	find	that	several	biological	
barriers	occlude	outer	membrane-localized	features	and	demonstrate	that	removal	of	these	
extracellular	components	by	either	chemical	or	genetic	means	increases	accessibility	to	the	
engineered	tags.	We	show	that	the	display	peptides	facilitate	specific	attachments	of	living	
	

33	

	

cyanobacteria	to	functionalized	beads	and	other	microbes,	a	technology	that	has	
applications	in	biomass	recovery	and	the	development	of	microbial	consortia.	

	

Figure	2.1.1:	Graphical	abstract	

Results	

	

Design	and	screening	of	surface	display	epitopes	

Surface	display	typically	utilizes	fusion	proteins	of	outer-layer	cell	components	to	
generate	surface	display	molecules	(Boder	and	Wittrup	2000;	Samuelson	et	al.	2002).	
SomA,	a	non-essential,	highly-expressed	(Rubin	et	al.	2015)	outer	membrane	porin,	was	an	
ideal	candidate	for	engineering	a	surface	display	system	(Umeda	et	al.	1996;	Hansel	et	al.	
1998).	The	HHpred	secondary	structure	prediction	program	(Söding	et	al.	2005)	predicted	
a	14	transmembrane-strand	β-barrel	structure	for	SomA	(Figure	2.1a)	which	differed	from	
previous	predictions	(Hansel	and	Tadros	1998),	both	of	which	informed	our	epitope	design	
(Figure	2.1.S1a,c).	To	minimize	the	possibility	of	misfolding	or	mistargeting	(Hagan	et	al.	
	

34	

	

2011;	Stapleton	et	al.	2015),	we	inserted	a	short	peptide	sequence	in	predicted	external	
loops	of	the	full-length	protein.	Potential	external	loops	were	selected	from	predicted	
structures	(Hansel	and	Tadros	1998)	and	we	generated	in-frame	insertions	to	create	a	
suite	of	genomically-integrated	constructs	(R1F-R5F)	with	the	FLAG	tag	driven	by	an	IPTG	
inducible	promoter	(Figure	2.1.2a,	Figure	2.1.S1d).		

	

Figure	2.1.2:	SomA	surface	display	design	and	efficacy		

(A)	HHPred	model	of	SomA	showing	predicted	extracellular	(EC),	outer	membrane	(OM),	
and	periplasmic	(PP)	regions.	Epitope	insertion	sites	are	labeled	and	flanking	amino	acid	
residues	listed	in	the	same	color.	(B)	Representative	fixed	cell	immunostaining	of	the	
indicated	strains	expressing	FLAG	tagged-SomA	at	the	respective	sites.	(scale	bars	=	2	µm).	
(C)	Live	R5F	cells	express	recombinant	SomA	(Western	Blot:	right),	but	exhibit	
impenetrant	staining	(left).	(D)	Representative	quantification	of	average	percent	of	cells	
labeled	from	fixed	(left)	or	live	(middle)	populations	paired	with	a	flow	plot	overlay	
comparing	R5FFixed	cells	(light	blue)	with	WTFixed	(purple).		
	
We	utilized	immunofluorescence	labeling	on	fixed	cells	to	screen	induced	SomA	
constructs	for	protein	expression	and	localization	(Figure	2.1.2b).	Recombinant	SomA	with	
inserts	in	target	regions	R1-R3	showed	very	low	levels	of	signal	while	SomA-R4F	appeared	
to	have	stochastic	signal,	with	only	a	small	number	of	strongly	fluorescent	cells	in	the	
population.	In	contrast,	SomA-R5F	demonstrated	strong	and	ubiquitous	signal	with	
relatively	minor	variation	between	cells	with	induction	and	strength	of	expression	further	
confirmed	by	western	blot	(Figure	2.1.2b,c	and	Figure	2.1.S2).		
	

35	

	

Extracellular	factors	occlude	outer	membrane	accessibility	in	living	cells	

We	repeated	the	immunolocalization	in	living	R5F	cells	to	determine	whether	the	
tag	remained	accessible	without	permeabilization.	We	observed	that	only	2%	of	living	cells	
were	labeled	as	compared	to	the	95%	of	cells	in	the	fixed	sample	(Figure	2.1.2d).	We	
hypothesized	that	external	cellular	components	might	be	sterically	occluding	the	epitope.	
While	cyanobacteria	are	gram-negative	like	in	that	they	possess	an	outer	membrane,	the	
outer	surfaces	of	most	cyanobacteria	–	including	S.	elongatus	–	are	poorly	characterized,	
and	can	differ	greatly	from	the	traditional	gram-negative	classification.	However,	some	
examples	of	extracellular	factors	that	can	be	found	on	the	surface	of	cyanobacteria	are	the	
lipopolysaccharide	layer	(LPS),	a	layer	of	long-chain	polysaccharides	covalently	linked	to	
molecules	in	the	outer	leaflet	of	the	outer	membrane	(Greenfield	and	Whitfield	2012),	and	
the	surface-layer	(S-layer),	a	paracrystalline	shell	of	a	single	glycoprotein	that	encapsulates	
the	cell	(Vaara	1982;	Schuster	and	Sleytr	2000;	Šmarda	et	al.	2002).	In	order	to	remove	
peripherally-associated	components	from	the	cell	surface,	we	adapted	an	S-layer	stripping	
protocol	which	utilizes	ethylenediaminetetraacetic	acid	(EDTA)	to	strip	away	metal	cations	
thought	to	stabilize	LPS	and	S-layers	(Sumper	et	al.	1990;	Messner	and	Sleytr	1992).	
Immunostaining	of	stripped	R5F	cells	(R5FEDTA)	showed	staining	penetrance	comparable	to	
fixed	cell	levels	without	the	loss	of	viability	(Figure	2.1.2a-b	and	Figure	2.1.S3),	supporting	
our	hypothesis	that	other	external	factors	were	occluding	the	tagged	SomA.		

	

36	

	

	

37	

LPS	extracellular	occlusion	of	epitope	

Figure	 2.1.3:	 Surface	 display	 epitope	 is	 occluded	 by	 EDTA-sensitive	 extracellular	
factor(s)		

(A)	Representative	images	of	immunolabeled,	induced,	live	R5F	cells	before	and	after	EDTA	
treatment	(R5FEDTA)	(scale	bars	=	5	μm).	(B)	Average	percentage	of	labeled	cells,	quantified	
via	flow	cytometry	where	n=3	independent-day	experiments	with	error	bars	representing	
the	s.d.	(P-values:	*<0.05,	**<0.01).	
Although	EDTA-based	stripping	protocols	can	remove	occluding	factors,	genetic	
ablation	of	the	underlying	extracellular	factors	could	constitutively	improve	epitope	
availability.	We	first	targeted	the	O-antigen	(OAg),	a	primary	component	of	the	LPS	
(Greenfield	and	Whitfield	2012)	in	 S.	elongatus	(Simkovsky	et	al.	2012)	by	ablating	OAg	
synthesis	and	transport	genes.	We	created	genetic	knockouts	(ΔwbdC,	Δwzm,	Δwzt)	of	the	
OAg	pathway	genes	in	wild	type	and	R5F	backgrounds.	No	significant	changes	in	viability	
or	morphology	were	observed	(Figure	2.1.S4).	We	reexamined	surface	epitope	accessibility	
in	living	cells	via	immunolocalization	and	flow	cytometry	(Figure	2.1.4a,b).	Expressing	the	
tagged	SomA	in	OAg-deficient	strains	resulted	in	a	moderate	increase	in	the	penetrance	of	
immunostaining	in	live	cells	(~10%)	(Figure	2.1.3b).	Interestingly,	numerous	labeled	cells	
had	a	patchy	appearance	(Figure	2.1.4a).	Incomplete	penetrance	and	variation	between	
	

	

experimental	replicates	indicates	the	OAg	may	contribute	to	FLAG	epitope	occlusion,	but	
there	are	likely	additional	occluding	factors.		

	

Figure	 2.1.4:	 Removal	 of	 OAg	 synthesis	 machinery	 and	 a	 putative	 S-layer	 protein	
improves	epitope	availability		

(A-C)	Live	cell	fluorescent	microscopy	and	flow	cytometry	of	the	R5F	surface	display	strain	
in	conjunction	with	gene	knockouts.	(A)	The	R5F	Δwzt	strain	shows	incomplete	penetrance	
and	often	patchy	staining	(orange	arrow).	(B)	Quantification	of	live	cell	immunostaining	
from	12	independent	experiments,	displaying	the	mean	value	with	error	bars	representing	
the	s.d.	and	significance	determined	by	a	two-tailed	student	t-test	(*	and	**	representing	P-
values	<0.05	and	<0.01	respectively).	(C)	The	R5F	ΔslpA	strain	shows	improved	levels	of	
staining,	but	penetrance	is	still	incomplete.	(D)	The	R5F	Δwzt	ΔslpA	strain	significantly	
improves	staining	of	the	population,	but	does	not	result	in	full	penetrance.	(E)	The	top	
three	LC-MS-MS	candidate	proteins	isolated	from	EDTA-stripped	cyanobacteria;	size	in	
kDa,	length	in	amino	acids	(see	Table	2.1.S1).	
We	examined	proteins	extracted	by	EDTA-stripping	for	candidate	surface	proteins	
that	might	be	occluding	the	peptide	tags.	Wild-type	and	R5F	cells	were	stripped,	released	
proteins	separated	by	SDS-PAGE,	and	the	most	abundant	bands	were	excised	for	LC-MS-MS	
	

Identification	and	deletion	of	extracellular	proteins	

38	

	

analysis	(Table	2.1.S1).	From	the	LC-MS-MS	data,	we	identified	three	top	candidates	based	
on	their	prevalence,	examined	their	predicted	homologs	by	BLAST	analysis	(Altschul	et	al.	
1990),	and	evaluated	their	candidacy	for	genetic	knockout	(Figure	2.1.4c-e	and	Table	
2.1.S1).	The	most	abundant	candidate	with	the	highest	confidence,	Synpcc7942_0422	
(hereafter:	slpA),	exhibited	homology	to	a	putative	S-layer	protein	from	Synechococcus	sp.	
PCC	7502	(WP_015167719).	Although	no	S-layer	has	been	conclusively	documented	in	S.	
elongatus,	metal	ion	chelators	are	known	to	disrupt	interactions	adhering	S-layer	proteins	
to	one	another	and	tethering	them	to	the	outer	surface	of	gram-negative	bacteria	(Messner	
and	Sleytr	1992).	
We	genetically	eliminated	each	of	the	top	candidate	proteins	to	examine	changes	in	
R5F	epitope	accessibility.	Knockout	lines	exhibited	normal	morphology	(Figure	2.1.S5),	and	
ΔslpA	significantly	improved	cell	live	immunolabeling	(Figure	2.1.4c).	Analysis	by	flow	
cytometry	showed	ΔslpA	staining	levels	consistently	ranging	from	20-35%	of	the	
population	(Figure	2.1.4b);	whereas	ΔsomB2	and	ΔslpB	staining	levels	were	unchanged.	As	
the	removal	of	 slpA	resulted	in	a	significant	and	uniform	increase	in	accessibility	of	the	
epitope,	it	is	likely	a	significant	factor	in	occluding	the	tagged	SomA.	However,	due	to	the	
incomplete	penetrance,	the	SlpA	protein	is	not	the	only	component	participating	in	this	
occlusion.		
	

Elimination	of	the	putative	s-layer	and	OAg	improves	surface	display	availability		

If	the	OAg	and	the	SlpA	protein	independently	contribute	to	epitope	occlusion,	
elimination	of	both	 wzt	and	 slpA	genes	would	additively	increase	surface	epitope	
	

39	

	

availability	in	the	R5F	strain.	We	generated	the	R5FΔwzt	ΔslpA	line	(Figure	2.1.S6),	and	
found	enhanced	immunostaining	relative	to	either	single	knockout	strains,	with	average	
cell	labeling	at	~45%	(Figure	2.1.4b,d).	Furthermore,	we	discovered	that	inducing	the	cells	
for	three	days	as	opposed	to	16	hours	significantly	increased	the	number	of	cells	labeled	
(Figure	2.1.S7).	We	therefore	opted	to	use	the	extended	induction	period	for	downstream	
experiments	(see	Materials	and	Methods).	Together	our	results	suggest	there	may	be	
numerous	extracellular	factors	on	the	surface	of	S.	elongatus	yet	to	be	defined,	all	of	which	
may	contribute	to	the	occlusion	of	outer	membrane	proteins.		

Surface	display	mediates	interactions	with	functionalized	abiotic	beads	

The	interaction	of	α-FLAG	antibodies	with	the	displayed	tag	demonstrates	that	the	
surface	epitope	is	accessible	to	diffusing	proteins,	and	therefore	other	soluble	molecules	
could	be	targeted	to	the	cyanobacterial	surface	in	an	analogous	manner.	Other	useful	
applications	for	surface	display,	such	as	the	specific	adhesion	of	cyanobacteria	to	large	
biotic	and	abiotic	surfaces,	requires	unimpeded	access	of	the	surface	epitope	as	the	target	
surfaces	are	unable	to	diffuse	through	any	remaining	external	components.	We	tested	
protein	A	coated	magnetic	beads	as	a	potential	binding	partner.	Protein	A	is	a	
Staphylococcus	aureus	surface	protein	that	binds	to	a	conserved	region	of	antibodies	(Fc),	
uninvolved	in	antigen	binding	(Sandor	and	Langone	1982).	The	beads	could	bind	to	
cyanobacteria	through	the	mediating	antibody	bound	to	the	surface	FLAG	tag.	In	initial	
tests,	we	utilized	R5FEDTA	cells	to	provide	a	uniformly	labeled	cell	population	(Figure	2.1.3).	
In	equal	mixes	of	beads	and	labeled	R5FEDTA	cells,	we	observed	small	cell/bead	aggregates	
	

40	

	

(Figure	2.1.5a).	These	aggregates	only	appeared	in	samples	that	were	both	induced	and	
labeled,	indicating	aggregate	formation	requires	expression	of	tagged	SomA	(Figure	2.1.S8).		
To	determine	the	number	of	cells	participating	in	these	aggregates,	an	excess	of	
beads	were	conjugated	to	FITC	and	then	incubated	with	R5FEDTA	cells.	These	mixtures	were	
analyzed	via	flow	cytometry.	Under	these	conditions,	R5FEDTA	cells	are	identifiable	by	
chlorophyll	autofluorescence	(P1)	while	isolated	FITC-conjugated	beads	solely	fluoresce	in	
the	green	(P2)	(Figure	2.1.5b).	Mixed	populations	where	the	surface	epitope	was	induced	
showed	almost	complete	depletion	of	the	unbound	cyanobacteria	population	and	the	
appearance	of	a	third	aggregate	population	(P3)	(Figure	2.1.5b),	indicative	of	physical	
association	of	cells	and	beads.	Since	multiple	cells	can	bind	to	a	single	bead	and	vice	versa,	
we	cannot	accurately	assess	the	bead	to	cell	ratio	in	a	given	aggregate.	However,	FACS	
analysis	indicated	~90%	of	the	cells	were	removed	from	the	unbound	cell	population	
(Figure	2.1.5b).	We	repeated	these	experiments	with	live	R5FΔwzt	ΔslpA	cells	(Figure	
2.1.5c)	and	found	that	this	strain	performed	similarly,	with	over	85%	of	the	cells	depleted	
from	the	unbound	population.		

	

41	

	

Figure	 2.1.5:	 Surface	 display	 can	 mediate	 interactions	 between	 cyanobacteria	 and	
abiotic	surfaces		

(A)	Representative	images	of	uninduced	and	induced	cells	with	protein	A	coated	beads	
(white	arrows).	(B)	Flow	cytometry	of	induced	or	uninduced	R5F	cyanobacteria	treated	
with	EDTA	(P1)	mixed	with	beads	(P2)	shows	the	formation	of	a	bead-cell	aggregate	
population	(P3)	in	the	induced	cell	state.	Bar	graphs	quantify	the	percentage	of	
cyanobacterial	counts	lost	from	the	P1	relative	to	control	cell	samples	run	without	beads;	
this	loss	is	interpreted	as	the	number	of	cells	participating	in	bead-cell	aggregates.	
Averages	of	n=3	experiments	displayed;	error	bars	represent	s.d.	P-values:	*<0.05,	
**<0.01).	(C)	Experiments	as	in	(B),	except	R5FΔwzt	ΔslpA	strains	were	used,	without		
	

	

42	

	

Figure	2.1.5	(cont’d)	

EDTA-stripping.	(D&E)	Magnet	assisted	pelleting	of	mixed	bead-cell	populations,	allows	for	
the	removal	of	bound	biomass.	(D)	A	representative	graph	of	counts	in	both	P1	(unbound	
cyanobacteria)	and	P3	(bead-cell	aggregates)	gates	in	mixed	R5FEDTA	flow	cytometry	
sample,	before	and	after	applying	the	magnet.	(E)	Quantifying	the	counts	from	P1	and	P3	
relative	to	control	samples	without	beads,	there	is	induction	dependent	depletion	of	
biomass.	(n=3	experiments,	error	bars	=	s.d.)	
	We	examined	the	capacity	for	these	functionalized	beads	to	facilitate	magnet-
assisted	cyanobacterial	biomass	recovery.	As	before,	induced	and	uninduced	living	
R5FΔwzt	ΔslpA	strains	were	mixed	with	protein	A	magnetic	beads	and	incubated	with	a	
mediating	α-FLAG	antibody.	A	magnetic	field	was	then	applied	to	recover	the	beads	from	
solution.	Comparing	supernatant	fractions	from	samples	and	controls	via	flow	cytometry,	
we	found	that	~75%	of	the	cyanobacteria	could	be	removed	by	magnet-assisted	recovery	
(Figure	2.1.5d),	approximately	all	cells	with	accessible	epitopes.	These	results	definitively	
demonstrate	the	ability	for	this	surface	display	system	to	mediate	binding	between	the	
engineered	cyanobacteria	and	functionalized	beads.		

Surface	display	mediates	binding	to	engineered	yeast	binding	partners	

Since	surface	display	allowed	effective	binding	of	cyanobacterial	cells	to	abiotic	
surfaces,	we	next	examined	the	potential	for	this	system	to	enable	binding	between	two	
distinct	microbial	species.	We	utilized	 Saccharomyces	 cerevisiae	as	a	partner	because	it	
possesses	a	well-characterized	surface	display	system	(Kondo	and	Ueda	2004)	that	can	be	
used	to	mediate	the	interaction.	The	EBY-100	strain	used	in	this	work	displays	the	protein	
A	domain	on	the	surface	of	its	cell	wall	via	expression	from	an	inducible	plasmid	system	
(Boder	and	Wittrup	2000)	and	is	an	ideal	strain	to	examine	the	feasibility	of	inter-species	
adhesion	of	cyanobacteria	while	allowing	for	the	continued	use	of	tested	antibody	pairings.		
	

43	

	

We	first	verified	expression	of	the	Aga2p-proteinA	fusion	in	induced	EBY-100	cells	
via	immunostaining	and	flow	cytometry,	and	observed	that	~75%	of	the	population	
appeared	to	express	the	fusion	protein	(Figure	2.1.S9a).	This	well-documented	inefficiency	
is	an	artifact	of	the	expression	vectors	and	dropout	media	selection	which	are	permissive	of	
plasmid	loss	(Andreu	and	Del	Olmo	2013).	Induced	S.	cerevisiae	were	mixed	with	an	excess	
of	labeled	R5FEDTA	cells.	Using	flow	cytometry,	we	detected	the	appearance	of	an	aggregate	
population	that	possessed	the	auto-fluorescent	properties	of	cyanobacterial	cells	with	the	
light	scattering	properties	of	yeast-sized	particles	or	larger	(Figure	2.1.6a).	The	formation	
of	aggregates	was	then	confirmed	via	fluorescent	microscopy	where	we	observed	
multicellular	aggregates	of	yeast	and	cyanobacteria	(Figure	2.1.6b).	These	experiments	
were	repeated	with	the	R5FΔwzt	ΔslpA	cells,	and	we	obtained	similar	results	with	yeast	
and	cyanobacteria	association	dependent	on	the	expression	of	the	surface	display	proteins	
(Figure	2.1.6c).	Furthermore,	we	determined	50%	of	yeast	were	bound	to	one	or	more	
cyanobacterial	cells	by	monitoring	the	loss	of	unbound	yeast	cells	(Figure	2.1.6a-b).	
Accounting	for	the	~25%	non-expressing	yeast,	~75%	of	the	potential	binding	yeast	have	
been	bound	into	an	aggregate.	Aggregate	formation	only	occurred	in	mixtures	containing	
labeled	induced	cyanobacteria	and	induced	yeast	(Figure	2.1.6b	and	Figure	2.1.S9b),	
indicating	adhesion	specificity.	Imaging	the	resultant	population,	we	observe	a	range	of	
interactions,	from	single	yeast-cyanobacterial	interactions	to	larger	cell	aggregates	(Figure	
2.1.6c	and	Figure	2.1.S9c).		

	

44	

	

Figure	 2.1.6:	 Surface	 display	 mediated	 adhesion	 of	 cyanobacteria	 to	 engineered	
yeast		

Mixing	induced	and	antibody	labeled	(A)	R5FEDTA	or	(B)	R5FΔwzt	ΔslpA	cyanobacteria	(P1)	
with	induced	engineered	yeast	(P2)	results	in	cyanobacteria-yeast	aggregate	formation	
(P3).	Average	%	of	unbound	yeast	in	n=3	experiments	displayed;	error	bars	represent	s.d.	
P-values:	*<0.05,	**<0.01)	(c)	Representative	images	of	varied	cyanobacteria/yeast	
aggregates	from	flow	cytometry	samples,	(scale	bar	=	5	µm).	(see	Figure	S10	for	additional	
images)	
We	have	developed	a	surface	display	system	in	 S.	 elongatus	by	expressing	a	
modified	endogenous	outer	membrane	porin	SomA.	In	order	for	heterologous	affinity	tags	
on	the	outer	membrane	of	 S.	 elongatus	to	be	accessible,	we	have	demonstrated	that	
multiple	occluding	factors	must	be	removed.	After	removal	of	these	factors,	we	
	

Discussion	

	

45	

	

conclusively	show	that	the	introduced	epitopes	are	displayed	by	living	cyanobacterial	cells	
and	are	largely	accessible	to	the	extracellular	media.	While	the	epitope	we	introduced	to	
the	surface	of	cyanobacteria	lacks	inherent	functionality,	it	allows	for	highly-specific	
engineered	interactions.	In	proof-of-concept	experiments,	we	have	shown	that	these	
surface	modifications	can	mediate	attachment	of	cyanobacteria	to	abiotic	surfaces	and	to	
other	microbes	engineered	with	a	complimentary	display	system.		Broadly,	this	work	
enables	the	progression	of	surface	display	technology	in	cyanobacteria	and	facilitates	
future	academic	and	industrial	applications	for	S.	elongatus.			
The	work	presented	here	is	an	advance	upon	previous	efforts	towards	developing	
cyanobacterial	surface	display,	in	part	because	we	document	extracellular	components	that	
occlude	surface	accessibility	and	which	may	account	for	the	incomplete	functionality	of	
previous	surface	engineering	efforts.	Chungjatupornchai	et	al.	have	conducted	the	most	
extensive	work	on	cyanobacterial	surface	display,	utilizing	fusions	of	ice	nucleation	protein	
(InP)	or	N-terminal	truncations	of	SomA	to	target	proteins	to	direct	them	to	the	surface	of	
S.	 elongatus	(Chungjatupornchai	and	Fa-Aroonsawat	2008;	Chungjatupornchai	and	Fa-
aroonsawat	2009;	Chungjatupornchai	et	al.	2011).	The	authors	concluded	that	the	majority	
of	surface	targeted	organophosphorus	hydrolase	(OPH)	(Chungjatupornchai	and	Fa-
Aroonsawat	2008;	Chungjatupornchai	et	al.	2011)	and	green	fluorescent	protein	(GFP)	
the	
(Chungjatupornchai	 and	 Fa-aroonsawat	 2009)	 remained	 sequestered	 within	
periplasm/cell	wall	or	were	otherwise	inaccessible	to	the	external	environment.	More	
recently,	Ferri	et	al.	(2015)	heterologously	expressed	the	 E.	 coli	antigen	43	(Ag43),	an	
autotransporter	protein	involved	in	mediating	cell-cell	contacts	and	biofilm	formation,	in	
Synechocystis	 sp.	6803.	The	protein	was	expressed	with	its	endogenous	N-terminus	or	
	

46	

	

introduced	

is	

truncated	and	fused	with	the	N-terminal	signal	peptide	of	SomA.	In	both	forms,	Ag43	was	
expressed,	but	did	not	promote	cell-cell	adhesion	or	biofilming.	Since	the	heterologous	
proteins	were	susceptible	to	proteinase	K-mediated	degradation,	it	is	possible	they	were	
localized	correctly,	but	non-functional	due	to	occlusion	of	Ag43	by	extracellular	
components,	such	as	the	previously-characterized	S-layer	(Trautner	and	Vermaas	2013).	
Alternatively,	as	protease	K	compromises	the	integrity	of	the	outer	membrane	under	a	
variety	of	conditions	(Besingi	and	Clark	2015),	it	is	also	possible	that	the	 E.	 coli	or	
truncated	SomA	signal	peptides	were	inefficient	in	targeting	Ag43	to	the	outer	membrane	
and	some	proportion	of	protein	remained	within	the	periplasmic	space.	These	deficiencies	
in	protein	outer	membrane	localization	and	activity	prompted	us	to	pursue	internal	tagging	
of	SomA,	since	truncations	of	the	β–barrel	porin	could	exacerbate	translocation	defects.	
Here	we	have	confirmed	outer	membrane	localization	and	accessibility	to	both	diffusible	
biomolecules	and	cell-sized	tethered	surfaces,	and	found	evidence	that	removal	of	LPS	
synthesis	machinery	and	a	putative	S-layer	protein	were	important	to	prevent	occlusion	of	
the	displayed	epitopes.	
It	should	be	noted	that	even	in	our	best	performing	strain	(R5FΔwzt	 ΔslpA),	the	
	
elimination	of	occluding	factors	was	incomplete;	the	epitope	of	~10-20%	of	cells	remained	
inaccessible.	It	is	likely	that	additional	components	remain	on	the	surface	of	cyanobacteria	
that	 additional	
that	 can	 partially	 obscure	
exopolysaccharides	are	being	produced	as	it	has	recently	been	shown	that	cyanobacteria	
may	utilize	a	variety	of	components	from	canonical	Wzy,	ABC-transporter,	and	Synthase-
dependent	extracellular	polymeric	substance	(EPS)	secretion	pathways	interchangeably	or	
identified	 a	 dominant	
redundantly	 (Pereira	et	al.	2015).	 Additionally,	 while	 we	
	

tags.	 One	 possibility	

47	

	

extracellular	protein	with	similarity	to	a	putative	S-layer	protein	in	Synechococcus	sp.	PCC	
7502,	conservation	of	S-layer	proteins	is	extremely	poor	and	it	is	therefore	difficult	to	
assess	if	alternative	S-layer	proteins	are	encoded	by	S.	elongatus.	Further	interrogation	of	S.	
elongatus’	surface	proteins	may	identify	additional	targets	to	reduce	external	obstructions.		
From	an	industrial	standpoint,	the	ability	to	program	materials	to	have	highly	
specific	interactions	with	cyanobacteria	has	many	possible	biotechnological	applications.	
We	demonstrate	here	that	surface	displaying	cyanobacteria	can	bind	magnetic	beads,	
allowing	for	the	majority	of	biomass	to	be	recovered	without	centrifugation,	chemical	
flocculants,	or	filtration.	Furthermore,	as	the	interaction	between	engineered	material	and	
cyanobacterial	cells	can	be	highly	specific,	it	may	be	possible	to	isolate	cyanobacteria	from	
non-specific	contaminants	that	are	a	major	problem	of	scaled	reactors	(Mata	et	al.	2010).	
While	our	exact	approach	is	not	economically	viable	for	large-scale	cultivation	due	to	the	
use	of	antibodies,	these	experiments	provide	evidence	that	surfaces	with	appropriate	
chemistry	could	be	used	for	novel	biomass	recovery	techniques,	possibly	ameliorating	
current	biomass	recovery	problems	(Grima	et	al.	2003).	A	more	thorough	exploration	of	
the	chemistries/size	of	peptides	that	can	be	displayed	on	the	surface	of	cyanobacteria	
would	expand	the	strategies	that	can	be	used,	including	to	those	that	do	not	rely	upon	
addition	of	antibodies	or	other	mediating	compounds.	Furthermore,	use	of	cyanobacterial	
surface	display	to	recover	biomass	would	be	more	economical	if	the	binding	agents	were	
inexpensive,	reusable,	and/or	the	interactions	reversible.	One	material	that	would	meet	
these	criteria	is	other	engineered	cells,	analogous	to	the	engineered	yeast	we	utilized	in	
these	experiments.	
	

48	

	

Cell-cell	attachment	is	a	common	theme	in	nature,	where	surface	ligands	and	
receptors	are	used	to	coordinate	cell	types,	from	multicellular	tissues	to	microbial	biofilms.	
We	show	that	complementary	surface	display	can	be	used	to	program	intercellular	and	
interspecies	interactions.	Beyond	biomass	recovery,	a	possible	application	for	these	
interactions	 would	 be	 to	 advance	 previously	 described	 synthetic	 co-cultures	 of	
cyanobacteria	and	heterotrophs,	where	S.	elongatus-derived	carbohydrates	support	diverse	
heterotrophs,	including	S.	cerevisiae	(Niederholtmeyer	et	al.	2010;	Ducat	et	al.	2012;	Smith	
and	Francis	2016;	Hays	et	al.	2017).	Direct	attachment	of	partners	in	designed	
communities	could	potentially	increase	metabolite	exchange	and	stabilize	synthetic	
commensal	relationships.	Future	strains	of	both	 S.	 elongatus	and	 S.	 cerevisiae	could	be	
specifically	modified	to	facilitate	facile	interactions,	including	the	insertion	of	larger	protein	
domains	into	SomA	(a	potential	mechanism	to	enzymatically	activate	the	surface	of	the	
cyanobacteria),	to	eliminate	the	need	for	antibodies	to	mediate	binding.	Because	of	the	
genetic	tools	within	these	model	organisms,	such	strains	might	also	be	useful	as	a	platform	
to	study	and	engineer	the	aggregation	dynamics	of	planktonic	microbes	into	structured,	
multi-species	assemblies.			
In	summary,	this	work	has	shown	the	development	of	a	surface	display	system	in	
Synechococcus	elongatus	PCC7942	allows	for	the	specific	adhesion	of	S.	elongatus	to	both	
beads	and	engineered	S.	cerevisiae	cells,	with	potential	implications	for	both	industrial	and	
academic	settings.	
	
	

49	

	

Materials	and	Methods	

S.	elongatus	cell	culture	and	strains		

Cyanobacterial	cultures	were	grown	in	baffled	flasks	in	a	Multitron	Pro	(Infors)	
incubator	under	constant	illumination	from	fluorescent	bulbs	(15W	Gro-Lux;	Sylvania;	~70	
µmol/m2/s),	constant	temperature	(32oC),	and	2%	CO2	supplementation	and	shaking	(125	
rpm).	Cultures	were	grown	in	BG-11	media	buffered	with	1	g/L	HEPES	(pH	8.3;	Sigma).	
Surface	display	constructs	were	obtained	by	cloning	 S.	 elongatus	SomA	into	a	NSIII	
integration	vector	(Niederholtmeyer	et	al.	2010).	The	epitope	tags	and	linker	coding	
sequence	were	inserted	using	primer	directed	mutagenesis.	Gene	knockout	plasmids	were	
constructed	by	cloning	~1	kilobase	fragments	from	upstream	and	downstream	of	target	
genes	into	a	plasmid	flanking	antibiotic	resistance	selectable	markers.	Plasmid	assembly	
was	accomplished	in	all	cases	through	standard	isothermal	assembly	techniques	(see	
Supplemental	Materials	for	Primer	List).	Transformations	were	performed	as	per	(Golden	
et	al.	1987)	and	transformants	were	selected	for	on	BG-11	agar	plates	with	appropriate	
antibiotics	(chloramphenicol	12.5	µg/ml,	kanamycin	16.7	μg/ml,	and	spectinomycin	100	
μg/ml)	and	verified	by	colony	PCR.	Modified	somA	expression	was	induced	with	the	
addition	of	1	mM	IPTG	(isopropyl-β-D-1-thiogalactopyranoside)	to	cultures	and	allowed	to	
incubate	overnight	(~16	hrs)	or	over	the	course	of	three	days	(~72	hrs),	as	indicated.	
Live	cells	were	pelleted,	washed,	and	blocked	(blocking	buffer:	5%	Bovine	Serum	
Albumin	(A9647;	Sigma)	in	PBS)	for	30	minutes	at	room	temperature,	before	incubation	
with	1:1000	dilution	of	primary	antibody,	α-FLAG	(DYKDDDDK)	tag	monoclonal	mouse	
	

Immunostaining	of	cyanobacterial	cells	

50	

	

IgG2b	(LT0420;	Lifetein)	for	1	hour	at	room	temperature.	Cells	were	then	washed	and	
resuspended	in	the	dark	in	blocking	buffer	with	secondary	antibody	at	1:1000	dilution,	as	
indicated;	goat	α	-mouse	IgG	Alexa	Fluor®	488	(35502;	Thermo)	for	basic	expression	
immunolabeling,	goat	α	-rabbit	IgG	Alexa	Fluor®	647	dye	(A-21236;	Thermo)	for	bead-cell	
adhesion	experiments,	or	rabbit	monoclonal	M87-3	anti-mouse	IgG1,	IgG2a,	IgG2b	H&L	
(ab125907;	abcam)	for	cyanobacteria-yeast	adhesion	experiments,	were	incubated	in	the	
dark	for	1	hour.	Stained	cells	were	washed	with	PBS	three	times	prior	to	binding	
experiments,	fluorescent	microscopy,	or	flow	cytometry.	Fixed	cells	(Figure	2.1.2)	were	
treated	as	above,	but	prior	to	blocking	were	resuspended	in	500	ul	of	100%	pre-chilled	
methanol	and	incubated	at	-20oC	for	10	minutes.	Cells	were	re-pelleted,	washed	with	1	ml	
of	PBS,	re-pelleted	and	resuspended	in	0.02	mg/ml	Lysozyme	(0663-10G;	Amresco)	in	Tris-
HCl	for	30	minutes	at	room	temperature	with	agitation.		

Cyanobacteria-bead	binding	assays	

Magnetic	protein	A/G	beads	(88802;	Thermofisher)	washed	and	resuspended	with	
1:1000	Fluoroscein-5-EX	succinimidyl	ester	(F6130;	Invitrogen)	for	1	hour.	Stained	beads	
were	washed	with	PBS	to	remove	excess	unconjugated	fluorophores	and	mixed	with	
immunolabeled	cyanobacterial	cells	(primary	and	goat	anti-rabbit	IgG	Alexa	Fluor®	647	
dye)	for	30	minutes	in	the	dark	at	room	temperature.	Mixtures	were	analyzed	by	flow	
cytometry	on	an	Accuri	C6	flow	cytometer	(653119;	BD	Biosciences)	with	a	minimum	
threshold	of	80,000	FSC-H,	or	by	microscopy	on	a	Zeiss	Axio	Observer.D1	(491911-0027-
000;	Zeiss).	
	

51	

	

Yeast	culturing	and	induction	conditions	

EBY-100	yeast	was	cultivated	as	per	(Boder	and	Wittrup	1997).	For	induction	of	
fusion	protein	expression,	yeast	from	overnight	cultures	were	back	diluted	into	SGCAA	
induction	media	(Boder	and	Wittrup	1997)	and	allowed	to	induce	for	4	hours	prior	to	
experimentation.	Induction	was	verified	via	immunolabeling	of	induced	cells	with	α-HA-
tag-Alexa	Fluor	488,	monoclonal	antibody	(M180-A48;	MBL)	which	targets	the	internal	HA	
tag	located	in	the	linker	region	bridging	the	Aga2a	domain	and	the	protein	A	domain	of	the	
surface	display	protein	(Boder	and	Wittrup	1997).	

Cyanobacteria-yeast	binding	assays	

Cyanobacteria	were	immunolabeled	as	described	above.	EBY-100	yeast	cells	were	
harvested,	incubated	in	blocking	buffer	for	30	minutes	at	room	temperature,	and	washed	in	
PBS.	These	cells	were	then	mixed	with	cyanobacteria,	(primary	and	secondary	(rabbit	
monoclonal	M87-3	anti-mouse	IgG1,	IgG2a,	IgG2b	H&L)),	at	room	temperature	for	30	
minutes	at	a	1:10	yeast:cyanobacteria	ratio.	Samples	were	analyzed	by	flow	cytometry	or	
microscopy	as	above.	

Surface	component	stripping	

This	protocol	was	adapted	from	(Sumper	et	al.	1990).	Briefly,	cells	were	harvested	
and	pelleted	from	culture	at	17,000	xg	for	10	minutes.	The	supernatant	was	removed	and	
the	cells	were	resuspended	in	192	mM	EDTA	(ethylenediaminetetraacetic	acid)	in	BG-11,	
and	incubated	at	room	temperature	for	30	minutes.	The	cells	were	then	pelleted	and	
resuspended	in	PBS	for	recovery.	EDTA	supernatant	was	retained	for	protein	analysis	by	
LC/MS/MS.	
	

52	

	

Western	blotting	

50	mL	of	culture	adjusted	to	0.3	OD750	was	lysed	with	a	benchtop	cell	disrupter	
(Constant	Systems	Ltd.);	three	passes	at	35	kPa	at	4oC.	Cell	lysates	were	concentrated	using	
15	ml	30,000	NMWL	Amicon	Centrifugal	Filter	Units	(UFC901008;	Millipore)	spun	at	3500	
xg	for	20	minutes	at	4oC.	Protein	lysate	was	separated	by	SDS-PAGE,	and	transferred	to	
PVDF	membrane	with	the	Trans-Blot	Turbo	Transfer	System	(1704155;	BioRad).	PDVF	
membranes	were	blocked,	and	probed	with	a	primary	antibody	1:1000,	α-FLAG	
(DYKDDDDK)	tag	monoclonal	mouse	IgG2b	(LT0420;	Lifetein)	overnight,	and	secondary	
antibody	1:1000,	chicken	α-mouse	IgG	secondary	antibody	HRP	Conjugate	(SA1-72021;	
Thermo)	for	4	hours.	The	membrane	was	imaged	using	a	Chemi	Doc	MP	(1708280;	
BioRad).	

Flow	cytometry	analysis	

Flow	cytometry	was	performed	using	a	BD	accuri	C6	flow	cytometer	(653119;	BD	
Biosciences).	The	particle	threshold	was	set	to	80,000	FSC-H	for	all	experiments	and	flow	
speed	was	optimized	for	fluorescent	measurements.	Gating	was	set	according	to	control	
population	samples	so	as	to	be	inclusive	across	experiments	and	exclusive	of	potential	
contaminants.	The	excitation	laser	used	for	all	measured	channels	was	the	20mW	488	nm	
Solid	State	Blue	Laser.	The	optical	filters	used	for	fluorescent	particles	were	FL1	533/30	
nm	(FITC/GFP)	for	FITC	and	AlexaFluor488	fluorescence,	and	FL3	>	670	nm	(PerCP,	PerCP-
CyTM5.5,	 PE-Cy7)	 for	 chlorophyll-α	 fluorescence.	 All	 flow	 cytometry	 experiments	
performed	 had	 multiple	 biological	 replicates	 as	
legends.	
Additionally,	for	each	biological	replicate,	three	technical	replicates	were	performed	(i.e.	
	

in	 the	 figure	

indicated	

53	

	

n=3	is	an	aggregate	of	9	sample	readings).	Combining	technical	replicates	averaged	80,000-
100,000	events	per	biological	replicate.	
	

	

	

54	

	

	

	

	

	

	

	

APPENDIX	

	

	

55	

	

56	

LC/MS/MS		

Appendix:	Chapter	2.1	Supplemental	Materials	

Standard	SDS-PAGE	of	EDTA-treated	cell	supernatants	was	conducted,	and	
dominant	bands	were	cut	from	4-20%	acrylamide	gradient	gels	(456-1094;	BioRad).	
Excised	bands	were	sent	to	the	Michigan	State	University	RTSF	Proteomics	Core	for	
downstream	processing	and	initial	analysis.	Gel	bands	were	digested	in-gel	according	to	
Shevchenko	et	al.	(1996)	with	modifications.	Briefly,	gel	bands	were	dehydrated	using	
100%	acetonitrile	and	incubated	with	10mM	dithiothreitol	in	100mM	ammonium	
bicarbonate,	pH~8,	at	56C	for	45min,	dehydrated	again	and	incubated	in	the	dark	with	
50mM	iodoacetamide	in	100mM	ammonium	bicarbonate	for	20min.	Gel	bands	were	then	
washed	with	ammonium	bicarbonate	and	dehydrated	again.	Sequencing	grade	modified	
typsin	was	prepared	to	0.01μg/μL	in	50mM	ammonium	bicarbonate	and	~50μL	of	this	was	
added	to	each	gel	band	so	that	the	gel	was	completely	submerged.	Bands	were	then	
incubated	at	37°C	overnight.	Peptides	were	extracted	from	the	gel	by	water	bath	sonication	
in	a	solution	of	60%ACN/1%TCA	and	vacuum	dried	to	~2μL.	Peptides	were	then	
resuspended	in	2%	acetonitrile/0.1%TFA	to	25μL.	From	this,	5μL	were	automatically	
injected	by	a	Thermo	EASYnLC	1000	onto	a	Thermo	Acclaim	PepMap	RSLC	0.075mm	x	
250mm	C18	column	and	eluted	over	16min	with	a	gradient	of	5%B	to	30%B	in	7min,	
ramping	to	100%B	at	8min	and	held	at	100%B	for	the	duration	of	the	run	or	eluted	over	
90min	with	a	gradient	of	5%B	to	30%B	in	79min,	ramping	to	100%B	at	80min	and	holding	
for	the	duration	of	the	run	(Buffer	A	=	99.9%	Water/0.1%	Formic	Acid,	Buffer	B	=	99.9%	
Acetonitrile/0.1%	Formic	Acid).	Eluted	peptides	were	sprayed	into	a	ThermoFisher	Q-
Exactive	mass	spectrometer	using	a	FlexSpray	spray	ion	source.	Survey	scans	were	taken	in	
	

	

the	Orbi	trap	(35000	resolution,	determined	at	m/z	200)	and	the	top	ten	ions	in	each	
survey	scan	are	then	subjected	to	automatic	higher	energy	collision	induced	dissociation	
(HCD)	with	fragment	spectra	acquired	at	17,500	resolution.	The	resulting	MS/MS	spectra	
are	converted	to	peak	lists	using	Mascot	Distiller,	v2.5.1	(www.matrixscience.com)	and	
searched	against	a	custom	database	containing	all	S.	elongatus	protein	sequences	available	
from	NCBI	(downloaded	from	www.ncbi.nlm.nih.gov,	2014-10-03)	and	appended	with	
common	laboratory	contaminants	(downloaded	from	www.thegpm.org,	cRAP	project)	
using	the	Mascot	searching	algorithm,	v	2.5.	The	Mascot	output	was	then	analyzed	using	
Scaffold,	 v4.3.4	 (www.proteomesoftware.com)	 to	 probabilistically	 validate	 protein	
identifications.	Assignments	validated	using	the	Scaffold	1%FDR	confidence	filter	are	
considered	true.	
Mascot	parameters	for	all	databases	were	as	follows:	allow	up	to	2	missed	tryptic	
sites;	fixed	modification	of	Carbamidomethyl	Cysteine;	variable	modification	of	Oxidation	
of	Methionine;	peptide	tolerance	of	+/-	5ppm;	MS/MS	tolerance	of	0.3	Da;	FDR	calculated	
using	randomized	database	search.	

 

	

 

57	

	

Figure	 2.1.S1:	 Comparisons	 of	 SomA	 models	 and	 NS3	 tagged	 SomA	 insertion	
construct	design		

(A)	A	figure	adapted	from	Mizuno	et	al.	(Umeda	et	al.	1996)	showing	the	predicted	
extracellular	(EC),	outer	membrane	(OM),	and	periplasmic	(PP)	regions	of	SomA	from	
Synechocystis	sp.	6803.	Regions	targeted	for	insertion	are	labeled	in	color	with	the	amino	
acid	residues	flanking	the	site	bolded	in	the	same	color.	R1	[290-291]	(red),	R2	[338/339]	
(green),	R3	[342/343]	(navy),	R4	[406/407]	(yellow),	R5	[480/481]	(cyan).	Phyre2-	(B)	
and	HHPred-	(C)	generated	models,	with	the	same	coloring	as	in	(A).	(D)	Neutral	site	3	
insertion	vector	containing	the	 somA	gene	cloned	directly	from	 Synechococcus	elongatus	
PCC	7942	labeled	with	target	insertion	sites.	

 

	

	

58	

	

59	

	

Figure	 2.1.S2:	 Additional	 representative	 immunolocalization	 of	 indicated	 tagged	
SomA	constructs	in	fixed	cells	

 (Top):	All	cells	(red:	chlorophyll)	are	induced	(+	1mM	IPTG);	the	insertion	site	of	the	
introduced	FLAG	tag	influences	the	expression	and	translocation	of	modified	SomA	
constructs	(green:	antibody	fluorescence).	Scale	bar	=	2	µm.	(Bottom):	Panels	as	above,	
except	where	only	the	antibody	fluorescence	channel	(Alexa	Fluor	488)	is	displayed	to	
improve	visualization	of	antibody	labeling.	Cells	are	either	induced	(+	1mM	IPTG)	or	
	
unininduced,	as	indicated.	Scale	bars	=	5	µm;	as	indicated.		
	

	

Figure	2.1.S3:	Cyanobacterial	cell	viability	following	EDTA	treatment		

Uninduced	and	induced	WT	and	NS3-SomA-R5F	cells	were	stripped	with	EDTA	to	disrupt	
extracellular	structures	and	inoculated	into	new	cultures.	Cells	remain	viable	following	
EDTA	treatment	(dotted	lines),	although	EDTA	may	cause	a	growth	lag,	as	indicated	by	the	
delayed	growth	in	the	first	24	hours.	

 

	

	

60	

	

	

Figure	2.1.S4:	O-antigen	knockout	construct	design	and	cell	morphology	of	resultant	
knockout	strains		

OAg	KO	constructs	were	designed	to	replace	the	target	gene	with	the	spectinomycin	
resistance	cassette	(left).	All	knockout	strains	exhibit	normal	cellular	morphology	(right).	
Scale	bar	=	5	µm.	

 

	

61	

Protein 

Grouping 

Ambiguity 

15+ kDa 

37+ kDa 

50+ kDa 

Band 

Band 

Band 

100+ 

kDa 

Band 

100% 

(15) 

100% 

(61) 

100% (6) 

73% (2) 

Compl

ex 

Sampl

e 

100% 

(5) 

100% (5) 

100% (2) 

TRUE 

100% (5) 

98% (1) 

100% 

(19) 

100% 

(13) 

-100% 

100% 

((0)) 

(22) 

-100% 

100% 

((0)) 

(18) 

100% (6) 

100% 

(31) 

-100% 

-100% 

((0)) 

((0)) 

19% 

(0) 

85% 

(4) 

100% 

(13) 

-100% 

((0)) 

-100% 

((0)) 

-100% 

((0)) 

100% 

(11) 

100% (9) 

100% (6) 

44% (2) 

100% (6) 

100% (6) 

100% (6) 

100% (6) 

14% (1) 

100% (8) 

-100% 

100% 

((0)) 

(8) 

-100% 

100% 

((0)) 

(5) 

-100% 

100% 

((0)) 

(12) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

((0)) 

100% (4) 

-100% 

100% 

((0)) 

(5) 

	

  

  

  

  

# 

1 

MS/MS View: 89 Proteins in 84 

Accession 

Molecular 

Clusters 

Number 

Weight 

hypothetical protein 

[Synechococcus elongatus] 

Chain B, X-ray Crystal Structure Of 

gi|499562623 

42 kDa 

2 

Phycocyanin From Synechococcus 

gi|459358742 

18 kDa 

elongatus Sp. Pcc 7942 

RecName: Full=C-phycocyanin 

alpha chain 

3 

iron deficiency induced protein A 

gi|130028 

17 kDa 

4 

[Synechococcus elongatus PCC 

gi|14331111 

39 kDa 

7942] 

5 

pili assembly chaperone 

[Synechococcus elongatus] 

gi|499696684 

15 kDa 

6 

porin [Synechococcus elongatus] 

gi|499561624 

58 kDa 

TRUE 

hypothetical protein 

Synpcc7942_1128 

[Synechococcus elongatus PCC 

7942] 

porin [Synechococcus elongatus 

PCC 7942] 

hypothetical protein 

[Synechococcus elongatus] 

7 

8 

9 

gi|81299937 

38 kDa 

  

gi|81300273 

57 kDa 

TRUE 

gi|499562638 

16 kDa 

  

RecName: Full=Allophycocyanin 

10 

100% (2) 

gi|129994 

17 kDa 

  

alpha chain 

Table	2.1.S1:	Complete	LC-MS-MS	peptide	reads	and	predicted	protein	matches	for	
EDTA-solubilized	proteins		

The	“Protein	Grouping	Ambiguity”	column	indicates	if	there	were	one	or	more	peptides	
that	could	be	mapped	to	more	than	one	discovered	protein.	Peptide	sequences	were	
evaluated	by	LC-MS-MS	(see	SI	Methods)	from	total	protein,	and	from	four	distinct,	
dominant	bands	(15,	37,	50,	100	kDa)	that	were	excised	from	SDS-PAGE	gels	following	
EDTA	treatment	of	living	S.	elongatus	cells	(Messner	et	al.	2008).	Percentiles	listed	in	the	
individual	band	columns	indicate	percent	certainty	of	the	proteins’	presence	in	each	
particular	band;	the	adjacent	number	in	parentheses	indicates	number	of	mapped	peptides,	
-100%	((0))	indicating	the	protein	was	not	detected.	The	three	top	candidates	(bold)	were	
identified	because	they	were	both	high	confidence	hits	in	the	total	protein	lysate	as	well	as	
identified	as	high-confidence	hits	in	excised	protein	bands	of	the	appropriate	size.	
Furthermore,	as	the	37+	kDa	band	was	dominant,	proteins	found	in	this	sample	were	given	
additional	prioritization.	
	

62	

	

	

Table	2.1.S1	(cont'd)	

11 

12 

SphX [Synechococcus elongatus 

PCC 7942] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

gi|496319 

36 kDa 

gi|499562813 

20 kDa 

13 

[Synechococcus elongatus PCC 

gi|24414819 

25 kDa 

7942] 

Chain B, X-ray Crystal Structure Of 

14 

Allophycocyanin From 

gi|459358658 

17 kDa 

15 

16 

17 

18 

19 

Synechococcus elongatus Pcc 7942 

alkaline phosphatase 

[Synechococcus elongatus] 

unknown [Synechococcus elongatus 

PCC 7942] 

ABC-type nitrate/nitrite transport 

system substrate-binding protein 

[Synechococcus elongatus PCC 

7942] 

hypothetical protein 

[Synechococcus elongatus] 

ANL06 [Synechococcus elongatus 

PCC 7942] 

C-terminal processing peptidase-2 

gi|499561694 

144 kDa 

gi|22002502 

17 kDa 

gi|81300048 

48 kDa 

gi|499563750 

20 kDa 

gi|47059644 

34 kDa 

20 

[Synechococcus elongatus PCC 

gi|81299523 

46 kDa 

7942] 

21 

22 

RecName: Full=Superoxide 

dismutase [Fe] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

gi|134647 

22 kDa 

gi|499563244 

43 kDa 

23 

[Synechococcus elongatus PCC 

gi|24414834 

19 kDa 

7942] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

24 

25 

26 

27 

gi|499563759 

37 kDa 

gi|499562724 

16 kDa 

gi|499697662 

9 kDa 

gi|499562523 

30 kDa 

Cluster of photosystem I reaction 

28 

center subunit II [Synechococcus 

gi|499697090 

16 kDa 

elongatus] (gi|499697090) 

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

63	

	
	

100% (3) 

100% (2) 

-100% 

-100% 

100% 

((0)) 

((0)) 

(4) 

100% (7) 

-100% 

((0)) 

39% (0) 

-100% 

100% 

((0)) 

(2) 

100% (4) 

-100% 

-100% 

-100% 

100% 

((0)) 

((0)) 

((0)) 

(2) 

100% (5) 

-100% 

((0)) 

98% (2) 

-100% 

100% 

((0)) 

(2) 

95% (1) 

100% (5) 

-100% 

-100% 

100% 

((0)) 

((0)) 

(2) 

100% (4) 

-100% 

-100% 

-100% 

100% 

((0)) 

((0)) 

((0)) 

(3) 

-100% 

-100% 

-100% 

-100% 

100% 

((0)) 

((0)) 

((0)) 

((0)) 

(3) 

100% (9) 

100% (2) 

-100% 

-100% 

-100% 

100% 

((0)) 

-100% 

((0)) 

((0)) 

((0)) 

(2) 

65% (0) 

-100% 

100% 

((0)) 

(2) 

-100% 

((0)) 

100% (9) 

-100% 

-100% 

((0)) 

((0)) 

13% 

(0) 

-100% 

-100% 

((0)) 

-100% 

((0)) 

((0)) 

100% (8) 

94% (2) 

-100% 

100% 

((0)) 

(4) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

100% (8) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

100% (3) 

100% (3) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

8% (1) 

98% (2) 

-100% 

100% 

((0)) 

(1) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

TRUE 

100% (6) 

99% (1) 

100% (1) 

((0)) 

-100% 

((0)) 

TRUE 

100% (2) 

0% ((0)) 

4% ((0)) 

0% ((0)) 

89% 

(1) 

	

Table	2.1.S1	(cont'd)	

photosystem I reaction center 

29 

subunit II [Synechococcus 

gi|499697090 

16 kDa 

TRUE 

100% (2) 

elongatus] 

photosystem I reaction center 

30 

subunit II [Synechococcus 

gi|499562072 

12 kDa 

TRUE 

35% ((0)) 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

elongatus] 

unknown [Synechococcus elongatus 

PCC 7942] 

orf134 [Synechococcus elongatus 

PCC 7942] 

hypothetical protein 

[Synechococcus elongatus] 

phosphohydrolase [Synechococcus 

elongatus] 

RecName: Full=Ycf48-like protein; 

Flags: Precursor 

pilin-like protein [Synechococcus 

elongatus PCC 7942] 

endo-1,4-beta-xylanase 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

purple acid phosphatase 

[Synechococcus elongatus] 

membrane protein [Synechococcus 

elongatus] 

rps12 [Synechococcus elongatus 

PCC 6301] 

hypothetical protein 

[Synechococcus elongatus] 

peptidylprolyl isomerase 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

RecName: Full=Protein RecA; 

AltName: Full=Recombinase A 

Cluster of 30 kD rod-rod linker 

gi|25019696 

21 kDa 

gi|454073 

14 kDa 

gi|499561785 

35 kDa 

gi|499562287 

94 kDa 

gi|108861977 

39 kDa 

gi|25019699 

20 kDa 

gi|499561686 

42 kDa 

gi|499562725 

17 kDa 

gi|499562279 

32 kDa 

gi|499562143 

65 kDa 

gi|1405430 

14 kDa 

gi|499561600 

22 kDa 

gi|499563276 

18 kDa 

gi|499563820 

17 kDa 

gi|123741845 

38 kDa 

  

  

  

  

  

  

  

  

  

  

  

  

  

  

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

89% 

(1) 

-100% 

((0)) 

4% ((0)) 

-100% 

-100% 

((0)) 

((0)) 

100% (5) 

100% (4) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

47% (0) 

100% (1) 

51% (0) 

23% (0) 

-100% 

-100% 

-100% 

-100% 

((0)) 

-100% 

((0)) 

100% (4) 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

100% (3) 

-100% 

-100% 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

100% (4) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

100% 

((0)) 

((0)) 

((0)) 

((0)) 

(1) 

-100% 

-100% 

-100% 

-100% 

100% 

((0)) 

((0)) 

99% (2) 

98% (0) 

((0)) 

-100% 

((0)) 

((0)) 

(1) 

44% (2) 

100% (2) 

100% (2) 

100% (3) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

((0)) 

((0)) 

48% (0) 

-100% 

100% 

((0)) 

(1) 

-100% 

((0)) 

99% 

(1) 

6% (0) 

((0)) 

4% 

((0)) 

-100% 

((0)) 

24% 

(0) 

100% 

(1) 

TRUE 

100% (4) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

46 

[Synechococcus elongatus PCC 

gi|142124 

30 kDa 

TRUE 

0% ((0)) 

0% ((0)) 

0% ((0)) 

0% ((0)) 

6301] (gi|142124) 

30 kD rod-rod linker 

47 

[Synechococcus elongatus PCC 

gi|142124 

30 kDa 

TRUE 

6301] 

64	

	
	

-100% 

-100% 

-100% 

-100% 

100% 

((0)) 

((0)) 

((0)) 

((0)) 

(1) 

	

Table	2.1.S1	(cont'd)	

Chain A, Solution Nmr Structure Of 

The Pbs Linker Domain Of 

Phycobilisome Rod Linker 

48 

Polypeptide From Synechococcus 

gi|315113184 

16 kDa 

TRUE 

elongatus, Northeast Structural 

Genomics Consortium Target 

Snr168a 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

unknown [Synechococcus elongatus 

PCC 7942] 

hypothetical protein 

[Synechococcus elongatus] 

membrane protein [Synechococcus 

elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

serine protease [Synechococcus 

elongatus] 

unknown [Synechococcus elongatus 

PCC 7942] 

hypothetical protein 

[Synechococcus elongatus] 

phycobilisome core component 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

gi|499563675 

18 kDa 

gi|499561723 

19 kDa 

gi|499697288 

19 kDa 

gi|499561621 

18 kDa 

gi|25019695 

52 kDa 

gi|499562653 

56 kDa 

gi|499563531 

33 kDa 

gi|499562226 

16 kDa 

gi|499561711 

14 kDa 

gi|499562467 

43 kDa 

gi|22002557 

25 kDa 

gi|499563369 

21 kDa 

61 

[Synechococcus elongatus PCC 

gi|2655259 

19 kDa 

7942] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

unknown [Synechococcus elongatus 

PCC 7942] 

hypothetical protein 

[Synechococcus elongatus] 

RecName: Full=30S ribosomal 

protein S8 

gi|499561689 

47 kDa 

gi|499563386 

19 kDa 

gi|25019697 

42 kDa 

gi|499562165 

21 kDa 

gi|123556368 

15 kDa 

62 

63 

64 

65 

66 

	

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

65	

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

2% 

((0)) 

100% (2) 

100% (2) 

100% (2) 

100% (2) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

100% (2) 

98% (1) 

99% (1) 

100% (2) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

42% 

(1) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

100% (1) 

100% (1) 

100% (2) 

100% (2) 

100% (2) 

100% (2) 

-100% 

-100% 

-100% 

-100% 

100% 

((0)) 

((0)) 

((0)) 

((0)) 

(1) 

-100% 

-100% 

-100% 

-100% 

100% 

((0)) 

((0)) 

((0)) 

((0)) 

(1) 

100% (1) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

99% (1) 

98% (1) 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

99% (2) 

99% (1) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

	

Table	2.1.S1	(cont'd)	

67 

68 

unknown [Synechococcus elongatus 

PCC 7942] 

hypothetical protein 

[Synechococcus elongatus] 

photosystem I reaction center 

gi|25019693 

21 kDa 

gi|499561641 

17 kDa 

69 

subunit III [Synechococcus 

gi|499561831 

17 kDa 

elongatus] 

hypothetical protein, partial 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

gi|499562457 

13 kDa 

gi|499562535 

16 kDa 

gi|499562873 

13 kDa 

ABC transporter substrate-binding 

protein [Synechococcus elongatus] 

gi|499562957 

28 kDa 

hypothetical protein 

[Synechococcus elongatus] 

hypothetical protein 

[Synechococcus elongatus] 

gi|499563689 

19 kDa 

gi|499696874 

13 kDa 

70 

71 

72 

73 

74 

75 

  

  

  

  

  

  

  

  

  

99% (1) 

99% (1) 

99% (1) 

99% (1) 

99% (1) 

99% (1) 

99% (1) 

99% (1) 

99% (1) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

-100% 

-100% 

-100% 

-100% 

((0)) 

((0)) 

((0)) 

((0)) 

Candidate Proteins 

Accession 

# 

BLAST Results 

Domains/F

amilies 

hypothetical protein [Synechococcus 

gi|499562

elongatus] (Synpcc7942_0422) 

623 

porin [Synechococcus elongatus] SomB2 

gi|499561

(Synpcc7942_1635) 

624 

hypothetical protein (Synpcc7942_1128) 

gi|812999

[Synechococcus elongatus PCC 7942] 

37 

Hypothetical Cyanobacterial Proteins, Putative 

S-Layer Protein (WP_015167719.1), Flagellar 

  

Hook Length control protein Flik (AIE74170.1),  

Porins, SomB [Synechococcus elongatus PCC 

Family, 

SLH Super 

6301], SomA [Synechococcus elongatus PCC 

6301] 

OprB 

Super 

Family 

outer membrane adhesin-like protein, 

Hemolysin-type calcium-binding region, 

hemolysin-type calcium-binding region, C-type 

DUF4214 

lectin domain-containing protein, cellulose 

binding protein  

 

	

 

66	

	

 

Figure	2.1.S5:	Potential	surface	layer	protein	vector	constructs	and	morphology	

Integration	vectors	targeting	the	endogenous	gene	via	~1	Kb	regions	of	homology	flanking	
the	gene	of	interest	were	used	to	delete	the	genes	and	replace	them	with	Kanamycin	
resistance	cassettes.	In	all	but	one	case,	we	were	able	to	delete	the	entire	gene	sequence.	
However,	the	GC	and	secondary	structure	rich	3’	end	of	 somB2	prevented	the	design	of	
suitable	primers	for	a	complete	deletion.	This	3’	fragment	was	frame-shifted	relative	to	the	
original	sequence	to	prevent	any	potential	production	of	the	peptide	fragment.	
Completeness	of	the	transformations	was	confirmed	by	genomic	PCR	and	the	resultant	
mutants	were	screened	for	gross	morphology	changes,	and	none	were	observed	(scale	bars	
=	5μm).	

 

	

	

67	

	

Figure	2.1.S6:	Additional	representative	images	of	live	R5F	wzt	slpA	cells	

(red:	chlorophyll	autofluorescence)	induced	for	16	hours	and	labeled	with	FLAG-tag	
specific	antibodies	(green).	Labeling	uniformity	is	slightly	improved	relative	to	single	
knockouts	(Figure	3A,D),	and	the	majority	of	cells	are	immunolabeled	to	some	extent	(scale	
bars	=	5	μm).	

 

	

	

68	

	

Figure	2.1.S7:	Longer	induction	of	SomA-R5F	increases	epitope	availability	

(Top)	Flow	cytometry	of	induced	(16	hours)	or	uninduced	R5FEDTA	or	R5FΔwzt	ΔslpA	cells	
(P1)	mixed	with	beads	(P2)	shows	the	formation	of	a	bead-cell	aggregate	population	(P3)	
in	the	induced	cell	state.	Bar	graphs	quantify	the	percentage	of	cyanobacterial	counts	lost	
from	the	P1	relative	to	control	cell	samples	run	without	beads,	interpreted	as	the	number	
of	cells	participating	in	bead-cell	aggregates.	Averages	of	n=3	experiments	displayed;	error	
bars	represent	s.d.	P-values:	*<0.05,	**<0.01).	(Bottom)	Experiments	as	in	above,	except	
cells	were	induced	for	a	total	of	72	hours	prior	to	mixing	with	beads.	Extending	the	
induction	of	cultures	increased	the	availability	of	the	epitope	in	live	cells,	as	evidenced	by	
increased	interaction	of	the	cells	with	the	beads.	
	
69	

 

	

	

 

Figure	2.1.S8:	Association	of	R5FEDTA	S.	elongatus	cells	with	magnetic	Protein	A	beads	
is	dependent	upon	IPTG-induced	expression	of	SomA-R5F	and	mediating	antibodies	

Cell/bead	mixtures	were	imaged	and	analyzed	for	the	formation	of	aggregates	under	a	
number	of	different	combinations	of	labeling	conditions,	demonstrating	the	necessity	for	
antibodies	as	well	as	the	expression	of	the	modified	SomA.	Neither	wild	type	cells	nor	
uninduced	R5F	cells	show	any	affinity	for	the	beads.	Aggregates	only	form	under	the	
specified	parameters.	Scale	bars	=	5	μm.	
	

	

70	

	

 

 

 

Figure	 2.1.S9:	 Induction	 of	 Protein	 A	 on	 the	 surface	 of	 EY100	 S.	 cerevisiae	 and	
antibody-mediated	adhesion	of	yeast	cells	to	R5F	S.	elongatus	

(A)	Representative	histogram	of	live	EY100	 S.	cerevisiae	cells;	either	uninduced	(red)	or	
induced	(blue)	to	express	and	display	HA-tagged	Protein	A.	Cells	are	labeled	with	anti-HA	
antibodies	(FITC-labeled).	Roughly	23%	of	the	induced	population	either	does	not	have	the	
expression	plasmid	or	is	not	expressing	the	protein	as	indicated	by	the	lack	of	antibody	
staining.	(B)	An	example	bar	graph	of	data	collected	from	a	flow	cytometry	experiment	
showing	the	lack	of	binding	between	cyanobacteria	and	yeast	cells	when	the	yeast	cells	are	
uninduced,	n=2	biological	replicates	with	3	technical	replicates	each.	Data	was	normalized	
to	cyanobacteria	control	samples	without	the	yeast	added.	(C-D)	Additional	examples	of	
R5F	 S.	 elongatus	cells	binding	to	EY100	 S.	 cerevisiae	in	(C)	small	cyanobacteria-yeast	
aggregates	or	(D)	larger	yeast	and	cyanobacteria	aggregates.	Scale	bars	as	indicated.	
	

71	

	

Primer Sequence (5'->3') 

Function 

5' Homology Target 

(for Isothermal 

Assembly) 

PCR somA from the Genome 

NSIII Vector Backbone 

PCR somA from the Genome 

NSIII Vector Backbone 

PCR NSIII Vector with homology 

to somA insert 

PCR NSIII Vector with homology 

to somA insert 

somA insert 

somA insert 

Protein Region 4 Site Insertion 

FLAG Tag and Linkers 

Protein Region 4 Site Insertion 

FLAG Tag and Linkers 

Protein Region 1 Site Insertion 

FLAG Tag and Linkers 

Protein Region 1 Site Insertion 

FLAG Tag and Linkers 

Protein Region 3 Site Insertion 

FLAG Tag and Linkers 

Protein Region 3 Site Insertion 

FLAG Tag and Linkers 

Protein Region 5 Site Insertion 

FLAG Tag and Linkers 

Protein Region 5 Site Insertion 

FLAG Tag and Linkers 

Protein Region 2 Site Insertion 

FLAG Tag and Linkers 

Protein Region 2 Site Insertion 

FLAG Tag and Linkers 

Sequencing Primer targeting 

within the somA gene 

Sequencing Primer targeting 

within the somA gene 

Sequencing Primer targeting the 

NSIII Site 

Sequencing Primer targeting the 

NSIII Site 

Sequencing Primer targeting the 

FLAG Tag Insert in somA 

N/A 

N/A 

N/A 

N/A 

N/A 

 

New 

# 

1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

	
	

Pseudonym 

Primer 

Direction 

F-somA 

Forward 

R-somA 

Reverse 

TCACACAAGGAGGAAAAACATATGTCTAGAATGAAACGCCTT

TTCTCGGCGCT 

CTTACGTGCCGATCGGATCCTACGCGGCCGCCTAGAAACGG

AAGACTGTTTTCAGGAAGG 

R-NS3 

Reverse 

TCTAGACATATGTTTTTCCTCCTTGTGTGAAATTGTTATCCGC 

R-NS3 

Forward 

CGGCCGCGTAGGATCCGATCGGCACGTAAGAG 

F-R4F 

Forward 

R-R4F 

Reverse 

F-R1F 

Forward 

R-R1F 

Reverse 

F-R3F 

Forward 

R-R3F 

Reverse 

F-R5F 

Forward 

R-R5F 

Reverse 

F-R2F 

Forward 

R-R2F 

Reverse 

ATCCGATTACAAAGATGATGACGATAAAGGAGGAGGATCCA

ACCTTGCTCTCTCCTTGGG 

TCCTTTATCGTCATCATCTTTGTAATCGGATCCACCGCCGTTG

CTGTTGTAGTTGGTAGC 

ATCCGATTACAAAGATGATGACGATAAAGGAGGAGGATCCA

AAGCCTTCAAGTATGTCGG 

CCTCCTTTATCGTCATCATCTTTGTAATCGGATCCACCGCCC

GTGTTCGCAGGGCCACTG 

GATCCGATTACAAAGATGATGACGATAAAGGAGGAGGATCC

GGTGTCTTTGGCTTCACCC 

TCCTTTATCGTCATCATCTTTGTAATCGGATCCACCGCCACCA

CTGTTGACAGTCGAGAC 

CCGATTACAAAGATGATGACGATAAAGGAGGAGGATCCTCT

GAAGACACTGGTTCCTTTG 

CCTCCTTTATCGTCATCATCTTTGTAATCGGATCCACCGCCC

GAACGGTTAGAAGCGCTG 

CCGATTACAAAGATGATGACGATAAAGGAGGAGGATCCACT

GTCAACAGTGGTGGTGTC 

CTCCTTTATCGTCATCATCTTTGTAATCGGATCCACCGCCCG

AGACATCAGCAGAGTTGC 

15 

F-somA-seq 

Forward 

TGATGTCTCCTTCACTGTTG 

16 

R-somA-seq 

Reverse 

GCAACTCGGAACTTGTTGTTATCAG 

17 

F-NS3-seq 

Forward 

GAGACGATGCGAACCTTCTCC 

18 

R-NS3-seq 

Reverse 

TCTGGCGACGGGACTACG 

19 

F-FLAGtag-

seq 

Forward 

GGATCCGATTACAAAGATGATGACGATAAAGGAGGA 

Table	2.1.S2:	Primer	List	

72	

	

Table	2.1.S2	(cont’d)	

Reverse 

TCCTCCTTTATCGTCATCATCTTTGTAATCGGATCC 

Forward 

CTATCAACTCACCACTGAAACTATTGCC 

Reverse 

ATGATTGTGGAAGGAGGCTCTGG 

Forward 

GGCCAGCGATCAATGGCAATTG 

Reverse 

CTAGCAGTTGTGCCAGTGCCG 

20 

21 

22 

23 

24 

R-FLAGtag-

seq 

F-wbdC-

upseq 

R-wbdC-

downseq 

F-wzm-

upseq 

R-wzm-

downseq 

25 

F-wzt-upseq 

Forward 

GTTGTTCGGCTGAGAGATTGGC 

Reverse 

GCAAACGCTTCGCAAGTTTAGG 

Forward 

ATGCGCTCACGCAACTGGCGTAGAAGAACAGCAAGGCCGCC

AATG 

Sequencing Primer targeting the 

FLAG Tag Insert in somA 

Sequencing Primer targeting 

upstream of the wbdC gene 

Sequencing Primer targeting 

downstream of the wbdC gene 

Sequencing Primer targeting 

upstream of the wzm gene 

Sequencing Primer targeting 

downstream of the wzm gene 

Sequencing Primer targeting 

upstream of the wzt gene 

Sequencing Primer targeting 

downstream of the wzt gene 

Amplifying Spectinomycin 

Resistance Cassette for OAg KO 

Plasmids 

Amplifying Spectinomycin 

N/A 

N/A 

N/A 

N/A 

N/A 

N/A 

N/A 

Universal Oag Upstream 

Homology (3’) 

26 

27 

28 

29 

30 

R-wzt-

downseq 

F-SpecR-

Oag 

R-SecR-

Oag 

F-pUC57-

Oag 

R-pUC57-

Oag 

33 

34 

35 

36 

37 

38 

F-DownH-

wzt 

R-DownH-

wzt 

F-UpH-

wbdC 

R-UpH-

wbdC 

F-DownH-

wbdC 

R-DownH-

wbdC 

	
	

Reverse 

TTATTTGCCGACTACCTTGGTGATCTCGCC 

Resistance Cassette for Oag KO 

N/A 

Forward 

CGCGTTGCTGGCGTTTTTC 

Reverse 

AGATCAAAGGATCTTCTTGAGATCCTTTTTTTCTG 

Plasmids 

Amplifying pUC57 Origin for Oag 

KO Plasmids 

Amplifying pUC57 Origin for Oag 

KO Plasmids 

N/A 

N/A 

31 

F-UpH-wzt 

Forward 

GATCTCAAGAAGATCCTTTGATCTGATCGATCACAGTAGCGT

Amplifying wzt Upstream 

TGATGGG 

Homology for KO Plasmid 

pUC57 Origin (3’) 

32 

R-UpH-wzt 

Reverse 

CCAGTTGCGTGAGCGCATAGAGGTATTGCCTATCACTCGCC 

Amplifying wzt Upstream 

Spectinomycin 

Homology for KO Plasmid 

Resistance Cassette (5’) 

Forward 

CACCAAGGTAGTCGGCAAATAAGCCATCATGAGAGAGTGGT

Amplifying wzt Downstream 

Spectinomycin 

TGTG 

Homology for KO Plasmid 

Resistance Cassette (3’) 

Reverse 

AAAAACGCCAGCAACGCGCCCACTACTGTAGGGCGAACTG 

Amplifying wzt Downstream 

Homology for KO Plasmid 

GATCTCAAGAAGATCCTTTGATCTGAGGGCTTTGATAATCGC

Amplifying wbdC Upstream 

CAGACTG 

Homology for KO Plasmid 

pUC57 Origin (5’) 

pUC57 Origin (3’) 

CCAGTTGCGTGAGCGCATCAAAATACTTGTGCTTGGTACTAA

Amplifying wbdC Upstream 

Universal Oag Upstream 

AATTGG 

Homology for KO Plasmid 

Homology (5’) 

CACCAAGGTAGTCGGCAAATAAAGCTGTTCGGCAGCTAACG

Amplifying wbdC Downstream 

Spectinomycin 

AG 

Homology for KO Plasmid 

Resistance Cassette (3’) 

Forward 

Reverse 

Forward 

Reverse 

AACGCCAGCAACGCGGCTGGAATAGTAGCCCCAGTTGAGG 

Amplifying wbdC Downstream 

Homology for KO Plasmid 

39 

F-UpH-wzm 

Forward 

GATCTCAAGAAGATCCTTTGATCTTGGGCTCGTGGATGCAGA

Amplifying wzm Upstream 

ACAG 

Homology for KO Plasmid 

pUC57 Origin (5’) 

pUC57 Origin (3’) 

40 

R-UpH-wzm 

Reverse 

CCAGTTGCGTGAGCGCATGGGCGGAGAGGGACAGCTG 

73	

Amplifying wzm Upstream 

Spectinomycin 

Homology for KO Plasmid 

Resistance Cassette (5’) 

	

Table	2.1.S2	(cont’d)	

CACCAAGGTAGTCGGCAAATAAAAATTTGCGACAGGGAGAC

Forward 

AGTC 

Reverse 

AAAAACGCCAGCAACGCGGTGGTTTCCAGCAAGTGTCC 

Amplifying wzm Downstream 

Spectinomycin 

Homology for KO Plasmid 

Resistance Cassette (3') 

Amplifying wzm Downstream 

Homology for KO Plasmid 

pUC57 Origin (5') 

CCTCGACGAGTGGCCTCTGACACATGCAGCTCGGAACTCAT

Amplifying pUC15 Origin for slpA 

slpA Downstream 

TGGCCGGACT 

KO 

Homology (3') 

TGGTGATTGACTACTACCTACCGTTACCCAACTTAATCGCCA

Amplifying pUC15 Origin for slpA 

slpA Upstream 

Forward 

Reverse 

GTTGAAACCAGCACCAGC 

KO 

Amplifying Kanamycin 

Homology (5') 

slpA Upstream 

Resistance Cassette for slpA KO 

Homology 

Forward 

GGATCCCTAGTACTCGCTAGGAAAGCCACGTTGTGTCTC 

Reverse 

GTTGTGATTTTTCAGTCTGGACCATCCCTAACTAAACAGTTGA

Amplifying Kanamycin 

slpA Downstream 

TTCG 

Resistance Cassette for slpA KO 

Homology 

47 

F-UpH-slpA 

Forward 

GCGATTAAGTTGGGTAACGGTAGGTAGTAGTCAATCACCAAC

Amplifying slpA Upstream 

GG 

Homology for KO Plasmid 

pUC15 Origin 

48 

R-UpH-slpA 

Reverse 

ACAACGTGGCTTTCCTAGCGAGTACTAGGGATCCCAG 

Amplifying slpA Upstream 

Kanamycin Resistance 

Homology for KO Plasmid 

Cassette 

CAACTGTTTAGTTAGGGATGGTCCAGACTGAAAAATCACAAC

Amplifying slpA Downstream 

Kanamycin Resistance 

Forward 

TGAG 

Forward 

CG 

Forward 

Reverse 

CG 

Reverse 

AGCTGCATGTGTCAGAGGCCACTCGTCGAGGCTAGC 

CAACTGTTTAGTTAGGGATGGTCTGAGATAGCTCGTACAGAC

Amplifying slpB Downstream 

Kanamycin Resistance 

Reverse 

GGCCAATGAGTTCCGCTGGCCACCCAAGTCCTGG 

TAAGTTTAAGTCTCTGAGAGTTAAATATTAGGAAAGCCACGT

Amplifying Kanamycin 

slpB Upstream 

TGTGTCTC 

Resistance Cassette for slpB KO 

Homology (3') 

CTGTACGAGCTATCTCAGACCATCCCTAACTAAACAGTTGATT

Amplifying Kanamycin 

slpB Downstream 

41 

42 

43 

44 

45 

46 

F-DownH-

wzm 

R-DownH-

wzm 

F-pUC15-

slpA 

R-pUC15-

slpA 

F-KanR-

slpA 

R-KanR-

slpA 

49 

50 

51 

52 

53 

54 

55 

56 

F-DownH-

slpA 

R-DownH-

slpA 

F-DownH-

slpB 

R-DownH-

slpB 

F-KanR-

slpB 

R-KanR-

slpB 

F-pUC15-

slpB 

R-pUC15-

slpB 

F-DownH-

somB2 

R-DownH-

somB2 

F-KanR-

somB2 

59 

60 

61 

	
	

Homology for KO Plasmid 

Cassette 

Amplifying slpA Downstream 

Homology for KO Plasmid 

pUC15 Origin (5') 

Homology for KO Plasmid 

Cassette (3') 

Amplifying slpB Downstream for 

KO Plasmid 

pUC15 Origin (5') 

Resistance Cassette for slpB KO 

Homology (5') 

Amplifying pUC15 Origin for slpB 

slpB Downstream 

KO 

Homology (3') 

Amplifying pUC15 Origin for slpB 

slpB Upstream 

KO 

Homology (5') 

Amplifying slpB Upstream 

Homology for KO Plasmid 

pUC15 Origin (3') 

Homology for KO Plasmid 

Cassette (5') 

Amplifying somB2 Downstream 

Kanamycin Resistance 

Homology for KO Plasmid 

Cassette (3') 

Amplifying somB2 Downstream 

Homology for KO Plasmid 

Amplifying Kanamycin 

pUC15 Origin (5') 

somB2 Upstream 

Homology (3') 

Forward 

ACTTGGGTGGCCAGCGGAACTCATTGGCCGGACT 

Reverse 

GCCGCGATCGATTGACAGTTGAAACCAGCACCAGC 

57 

F-UpH-slpB 

Forward 

GGTGCTGGTTTCAACTGTCAATCGATCGCGGCATCTAC 

58 

R-UpH-slpB 

Reverse 

TAGAAGAG 

ACAACGTGGCTTTCCTAATATTTAACTCTCAGAGACTTAAACT

Amplifying slpB Upstream 

Kanamycin Resistance 

Forward 

CAACTGTTTAGTTAGGGATGGGAACTCAATCCTCACACCGGG 

Reverse 

GGCCAATGAGTTCCGTGCCGTTAGGATAGGGGGC 

Forward 

TCGTGACAGCATAGAGGAGGAAAGCCACGTTGTGTCTC 

Resistance Cassette for somB2 

KO 

74	

	

Table	2.1.S2	(cont’d)	

62 

63 

64 

65 

66 

R-KanR-

somB2 

F-pUC15-

somB2 

R-pUC15-

somB2 

F-UpH-

somB2 

R-UpH-

somB2 

Reverse 

GGTGTGAGGATTGAGTTCCCATCCCTAACTAAACAGTTGATT

CG 

Forward 

CCCTATCCTAACGGCACGGAACTCATTGGCCGGACT 

Reverse 

GCAACTCATTCGGTAGATTTCAGTTGAAACCAGCACCAGC 

Forward 

GGTGCTGGTTTCAACTGAAATCTACCGAATGAGTTGCCCT 

Reverse 

ACAACGTGGCTTTCCTCCTCTATGCTGTCACGAATGCA 

67 

F-slpA-seq 

Forward 

GAGCGGATTGGCCAACGCATTC 

68 

R-slpA-seq 

Reverse 

GCGTACGACCCATCACAACAAAACC 

69 

70 

F-somB2-

seq 

R-somB2-

seq 

Forward 

CTAGAACAAGAACTCGGTCTTGAGTACTGGC 

Reverse 

ATGAAAAATCTGTTCAAGGTCATGTTGGCTG 

71 

F-slpB-seq 

Forward 

ATGGCTAATCTGAAGATTACTTCTGCTCAGCAG 

72 

R-slpB-seq 

Reverse 

TCAAACAATGGCAAGACTGGTGAGGTCG 

Amplifying Kanamycin 

Resistance Cassette for somB2 

KO 

somB2 Downstream 

Homology (5') 

Amplifying pUC15 Origin for 

somB2 Downstream 

somB2 KO 

Homology (3') 

Amplifying pUC15 Origin for 

somB2 Upstream 

somB2 KO 

Homology (5') 

Amplifying somB2 Upstream 

Homology for KO Plasmid 

pUC15 Origin (3') 

Amplifying somB2 Upstream 

Kanamycin Resistance 

Homology for KO Plasmid 

Cassette (5') 

Sequencing Primer targeting 

within the slpA gene 

Sequencing Primer targeting 

within the slpA gene 

Sequencing Primer targeting 

within the somB2 gene 

Sequencing Primer targeting 

within the somB2 gene 

Sequencing Primer targeting 

within the slpB gene 

Sequencing Primer targeting 

within the slpB gene 

N/A 

N/A 

N/A 

N/A 

N/A 

N/A 

For primers involved in isothermal assembly steps: Bold = (5' Homology);  Underline= (3' Binding Region) 

  

75	

	

	

CHAPTER	2.2:	ENGINEERING	OF	PAIRED	CYANOBACTERIAL	AND	HETEROTROPHIC	

SURFACE	DISPLAY	FOR	INTERSPECIES	ADHESION	

The	work	in	this	chapter	is	unpublished.	
	

76	

		
	

	

	

Introduction	

Microbial	consortia	are	ubiquitous	in	nature.	In	fact,	most	microbial	species	have	
become	so	specialized	they	cannot	function	outside	of	consortia	(Little	et	al.	2008).	These	
communities	are	often	quite	stable	when	perturbed	(Coyte	et	al.	2015)	and	play	multiple	
roles	in	the	environment	due	to	the	diversity	of	their	constituents	(Kato	et	al.	2005;	Coyte	
et	al.	2015;	Trivedi	et	al.	2016).	Both	enhanced	culture	stability	and	the	potential	for	
multifunctional	cultures	are	highly	desirable	features	in	industrial	settings.	Thus,	there	has	
been	increasing	interest	in	developing	designer	synthetic	consortia	for	the	targeted	
production	of	valuable	biological	compounds	(Sabra	et	al.	2010;	Ortiz-Marquez	et	al.	2013);	
yet,	there	are	few	examples	of	applied	artificial	consortia	at	scale	(e.g.,	the	production	of	
vitamin	C	(Ye	et	al.	2014)).	This	deficit	may	be	at	least	partially	explained	by	the	absence	of	
parameters	that	can	be	used	to	guide	the	development	of	industrially	viable	consortia.	In	
order	to	identify	the	fundamental	design	principles,	a	greater	understanding	of	how	
microbial	species	behave	in	both	natural	and	artificial	communities	is	needed.		
A	number	of	synthetic	consortia	have	been	engineered	to	explore	different	types	of	
microbial	interactions	and	how	they	influence	species	dynamics	within	natural	consortia	
(Brenner	et	al.	2008;	Kerner	et	al.	2012;	Ding	et	al.	2016;	Hays	et	al.	2017;	Tecon	and	Or	
2017).	A	significant	portion	of	this	foundational	research	has	been	performed	with	single-
species	auxotrophs	that	rely	on	artificially	induced	cross-feeding	scenarios	(Kerner	et	al.	
2012;	Tecon	and	Or	2017).	While	these	consortia	are	easily	tuned,	controlled,	and	
measured,	they	are	less	realistic	and	limit	the	kinds	of	conclusions	that	can	be	made	
regarding	community	dynamics.	One	of	the	ways	in	which	researchers	are	attempting	to	
generate	artificial	consortia	more	reflective	of	natural	communities	is	through	the	selection	
	

77	

	

of	 community	 members	 with	 different	 trophic	
lifestyles	 (e.g.,	 phototrophic	 vs.	
chemotrophic	vs.	heterotrophic).	Recently,	work	from	our	lab	and	others	has	explored	
artificial	consortia	utilizing	a	strain	of	the	cyanobacterium	 Synechococcus	 elongatus	PCC	
7942	engineered	to	excrete	endogenously	synthesized	sucrose	(Ducat	et	al.	2012)	in	
combination	with	different	heterotrophic	microbes	(Smith	and	Francis	2016;	Hays	et	al.	
2017;	Löwe	et	al.	2017a;	Weiss	et	al.	2017).	These	studies	have	demonstrated	that	this	
engineered	strain	of	cyanobacteria	has	the	ability	to	stably	support	a	variety	of	
heterotrophs,	operating	as	a	flexible	photosynthetic	co-culture	platform.		
With	sucrose	as	the	exchanged	metabolite,	any	species	that	either	naturally	utilizes	
sucrose	or	can	be	engineered	to	do	so	(e.g.,	Pseudomonas	putida	(Löwe	et	al.	2017b))	could	
potentially	be	paired	with	the	 S.	 elongatus	CscB	in	a	synthetic	consortium.	However,	it	
leaves	the	consortia	vulnerable	to	invasion	by	other	microbes	also	capable	of	metabolizing	
sucrose	(e.g.,	Stenotrophomonas	maltophilia	(Weiss	et	al.	2017)).	Additionally,	the	diffusion	
of	sucrose	results	in	individual	heterotrophs	experiencing	relatively	low	concentrations	of	
sucrose,	which	can	be	problematic	when	co-cultivating	model	heterotrophs	adapted	to	
growing	in	high-carbon	environments	(Kinoshita	1972).	To	address	these	limitations	I	
looked	to	naturally	cooperative	species	and	consortia	for	a	method	that	could	be	used	to	
potentially	stabilize	artificial	consortia.			
One	common	mechanism	used	to	maintain	selectivity	between	cooperating	organisms	
is	spatial	organization.	This	can	be	seen	in	a	number	of	natural	systems	including	lichen	
communities	(Pankratov	et	al.	2017),	marine	corals	(Kvennefors	et	al.	2017),	biofilms	
(Stoodley	et	al.	2002),	and	many	more	(Kim	et	al.	2008).	Could	physical	association	
between	the	engineered	cyanobacteria	and	their	heterotrophic	partners	increase	the	
	

78	

	

efficiency	of	carbon	transfer	and	decrease	risk	of	co-culture	contamination?	In	Chapter	2.1,	
I	showed	that	an	engineered	recombinant	SomA	protein	with	an	internal	FLAG-tag	epitope	
could	be	displayed	in	S.	elongatus,	and	demonstrated	that	this	surface	tag	could	be	used	to	
adhere	the	cyanobacteria	cells	to	functionalized	magnetic	agarose	beads	or	Saccharomyces	
cerevisiae.	While	this	system	did	demonstrate	the	potential	of	biologically	encoded	
adhesion	to	engineered	surface	substrates,	one	of	the	key	caveats	to	this	work	was	the	
required	addition	of	monoclonal	antibodies	to	mediate	the	binding	of	this	cyanobacterial	
strain.	In	this	chapter,	I	describe	preliminary	data	on	the	diversification	of	the	
cyanobacterial	surface	display	system	developed	in	Chapter	2.1	with	the	eventual	goal	of	
specific	interspecies	adhesion	in	co-culture.		
Specifically,	I	explored	the	capacity	of	the	SpyTag/SpyCatcher	and	StrepII/Streptavidin	
protein	domains	to	mediate	direct	attachment	between	a	modified	cyanobacterium	and	a	
heterotrophic	partner	species.	The	SpyTag/SpyCatcher	system	was	developed	from	an	
immunoglobulin-like	collagen	adhesion	domain	(CnaB2)	of	the	fibronectin	binding	protein	
(FbaB)	in	the	bacterium	Streptococcus	pyogenes	(Zakeri	et	al.	2012).	This	domain	was	one	
of	many	extracellular	proteins	stabilized	by	intramolecular	isopeptide	bonds	(Kang	and	
Baker	2011).	Zakeri	et	al.	split	this	domain	into	cognate	SpyTag	and	SpyCatcher	peptide	
sequences.	The	SpyTag	component	is	13	amino	acids	in	length	and	contains	the	reactive	
Asp117	residue	while	the	SpyCatcher	peptide	is	116	amino	acids	in	length,	constituting	the	
rest	of	the	CnaB2	domain	and	contains	the	Lys31	that	forms	the	isopeptide	bond	with	the	
SpyTag	 Asp117.	 Recombinant	 proteins	 containing	 the	 corresponding	 SpyTag	 and	
SpyCatcher	sequences	covalently	bond	to	one	another	at	a	range	of	temperatures	and	pH,	
enabling	stable	association	of	independently	encoded	and	expressed	proteins	(Zakeri	et	al.	
	

79	

	
2012;	Reddington	and	Howarth	2015).	This	split	domain	has	since	been	used	for	a	number	
of	applications,	including	enzyme	stabilization,	immune	system	stimulation,	and	protein	
scaffolding	(Reddington	and	Howarth	2015).		
The	Strep-tag	II	is	a	modified	version	of	the	original	9	amino	acid	affinity	tag	“Strep	tag”	
which	has	a	high	intrinsic	binding	affinity	for	the	protein	streptavidin.	Unlike	the	Strep-tag,	
Strep-tag	II	maintains	streptavidin	binding	capabilities	even	if	it	is	not	localized	to	the	C-
terminus	of	recombinant	proteins	(Schmidt	and	Skerra	1994).	However,	the	changes	to	the	
Strep-tag	II	amino	acid	sequence	also	resulted	in	a	lowered	affinity	for	streptavidin.	This	
was	later	rectified	through	the	generation	of	two	streptavidin	core	mutants	with	improved	
binding	affinity	(Voss	and	Skerra	1997)(Korndorfer	and	Skerra	2002).		In	this	work,	we	are	
utilizing	the	Streptavidin	Core	Mutant	2,	herein	referred	to	as	StrepCoreMut2	(Voss	and	
Skerra	1997),	as	the	basis	for	our	Strep-tag	II	affinity	system.	Both	the	StrepCoreMut2	
domain	and	SpyCatcher	domain	will	be	engineered	to	display	on	the	surface	of	the	
partnering	heterotrophs.	Here	we	generate	modified	strains	of	S.	elongatus,	E.	coli	W	cscR,	
and	 S.	 cerevisiae	EBY100	capable	of	expressing	and	translocating	recombinant	surface	
display	proteins	with	both	SpyTag/SpyCatcher	and	Strep-tag	II/Streptavidin	protein	
domains	to	their	surface.			

Results	

Design	of	surface	display	strains	

As	described	in	Chapter	2.1,	the	cyanobacterial	surface	display	system	requires	
insertion	of	the	displayed	moiety	into	a	presumptive	extracellular	loop	of	SomA	(Chapter	
2.1)	(Fedeson	and	Ducat	2016).	I	therefore	cloned	the	short	SpyTag	(13aa)	or	Strep-tag	II	
	

80	

	

(9	aa)	peptide	sequences	into	the	predicted	external	loop	(R5)	of	SomA	within	an	IPTG	
inducible	NSIII	vector	(Chapter	2.1);	naming	the	resulting	constructs	NSIII-SomA-
R5SpyTag	and	NSIII-SomA-R5StrepII,	respectively	(Figure	2.2.1A).	The	NSIII-SomA-
R5StreptagII	construct	simply	replaces	the	FLAG	tag	epitope	(described	in	chapter	2.1)	
with	the	Strep-tag	II	epitope	[NWSHPQFEK].	This	insert	is	flanked	on	either	side	by	4	aa	
linker	regions	[GGGS]	(Figure	2.2.1A).	The	NSIII-SomA-R5SpyTag	construct	contains	two	
epitope	tags,	the	SpyTag	[AHIVMVDAYKPTK]	and	the	FLAG	Tag	[DYKDDDDK],	both	are	
flanked	by	the	same	linker	sequence	[GGGS]	(Figure	2.2.1A).	The	FLAG	Tag	was	included	in	
this	construct	to	provide	an	antibody	epitope	for	validating	the	surface	localization	of	this	
protein.	Here	both	the	SpyTag	and	Strep-tag	II	epitopes	are	generically	referred	to	as	ligand	
domains.	Subsequent	transformation	and	chromosomal	integration	of	these	constructs	in	
wild-type	S.	elongatus	generated	the	R5SpyTag	and	R5StrepTagII	strains.	Transcription	of	
these	recombinant	genes	are	controlled	by	the	 trc	promoter	utilized	in	Chapter	2.1	and	
expression	can	be	induced	with	the	addition	of	IPTG	as	previously	described	(Fedeson	and	
Ducat	2016).	Inducing	both	of	these	strains	overnight	with	1	mM	IPTG,	live	cells	were	
harvested	from	culture	the	following	day,	stripped	of	extracellular	components	as	
previously	described	with	an	EDTA-based	stripping	buffer	(Fedeson	and	Ducat	2016),	and	
then	immunostained	with	primary	and	secondary	antibodies.	Fluorescent	microscopy	of	
these	samples	showed	fluorescent	antibody	localization,	indicating	that	the	antibodies	
were	able	to	interact	with	the	inserted	peptide	sequences	(Figure	2.2.1B).		
Both	S.	cerevisiae	and	E.	coli	have	extensive	histories	with	surface	display	systems	
(Pepper	et	al.	2008;	van	Bloois	et	al.	2011)	that	we	could	modify	for	our	purposes,	and	can	
also	be	stably	co-cultivated	with	 S.	elongatus	CscB	(Ducat	et	al.	2012;	Hays	et	al.	2017).	
	

81	

	

Therefore,	we	adapted	strains	of	these	model	heterotrophs	to	display	compatible	ligand	
pairs	to	the	cyanobacterial	R5SpyTag	and	R5StrepTagII	strain	surface	tags.		

Figure	2.2.1:	S.	elongatus	SpyTag	and	Strep-tag	II	surface	display	constructs	

(A)	Transcription	of	both	the	SomA-R5StrepTagII	(Top)	and	SomA-R5SpyTag	(Bottom)	
genes	is	driven	by	the	IPTG	inducible	trc-promoter	from	the	neutral	site	III	of	the	 S.	
elongatus	chromosome	(Chapter	2.1).	Each	tag	is	flanked	by	a	set	of	GlyGlyGlySer	linkers.	
(B)	Live	cell	immunostaining	of	the	NSIII-SomA-R5SpyTag	with	α-FLAG	tag	monoclonal	
mouse	IgG2b	(LT0420;	Lifetein)	(Top)	and	NSIII-SomA-R5StrepTagII	(anti-Strep-tagII)	
(Bottom)	show	that	the	integrated	NSIII-SomA	cassettes	are	inducible	and	that	the	protein	
is	translocated	to	the	cell	surface.	
	

	

82	

	

To	display	appropriate	binding	domains	on	S.	cerevisiae,	we	modified	the	published	
yeast	display	plasmid	(pETCON-Aga2p-proteinA;	(Boder	and	Wittrup	2000;	Fedeson	and	
Ducat	2016)),	replacing	the	protein	A	domain	with	sequence	for	Streptavidin	Mut2	or	
SpyCatcher	(See	constructs	pETCON-Aga2p-StreptavidinCoreMut2	and	pETCON-Aga2p-
SpyCatcher;	Figure	2.2.2A).	Additionally,	both	protein	sequences	contain	a	c-terminal	c-
Myc	peptide	tag	for	the	purpose	of	immunohistochemistry	and	western	blotting.	These	
constructs	were	introduced	into	chemically	competent	S.	cerevisiae	EBY100	cells	to	create	
StrepCoreMut2	and	SpyCatcher	yeast	surface	display	strains.	Protein	expression	from	these	
plasmids	was	induced	by	exposing	the	cells	to	SGCCA	medium	for	a	duration	of	4	hours	as	
was	done	in	(Chapter	2.1	Methods)	(Fedeson	and	Ducat	2016).		These	cells	were	then	
immunostained	with	primary	α-cMyc	antibodies	and	AlexaFluor	448	conjugated	secondary	
antibodies	(Chapter	2.1	Methods)	(Fedeson	and	Ducat	2016).	Imaging	of	these	strains	
revealed	that	both	constructs	display	the	recombinant	proteins	on	their	surface	(Figure	
2.2.2B).		

	

83	

	

84	

	

Figure	2.2.2:	S.	cerevisiae	SpyCatcher	and	StrepCoreMut2	surface	display	constructs	

(A)	Sequence	design	for	the	Aga2p-StrepCoreMut2	surface	display	protein	(Top)	and	the	
Aga2p-SpyCatcher	surface	display	protein	(Bottom)	are	both	derivatives	of	the	pETCON-
in	 Chapter	 2.1.	 (B)	
Aga2P-proteinA	 expression	 plasmid	 utilized	 and	 described	
Immunostaining	of	uninduced	and	induced	 S.	cerevisiae	cells	containing	these	constructs	
with	anti-cMyc	monocolonal	antibodies	and	AlexaFluor488	conjugated	anti-mouse	
secondary	antibodies	(as	per	Chapter	2.1	methods)	show	that	both	the	SpyCatcher	and	
StrepCoreMut2	constructs	are	successfully	displayed	on	the	surface	of	the	yeast	cells.		
	

	

For	display	of	compatible	binding	domains	on	the	surface	of	E.	coli,	we	explored	two	
conventional	systems.	The	first	system	we	designed	uses	a	combination	of	native	 E.	coli	
protein	sequences,	specifically	a	lippopolyprotein	signal	peptide	fragment	(Lpp’),	and	a	
truncated	outer	membrane	protein	A	(OmpA),	the	N-terminus	of	which	is	fused	to	the	
protein	of	interest	that	is	to	be	displayed	(van	Bloois	et	al.	2011).	Using	this	framework,	we	
generated	two	constructs	that	would	express	Lpp’-OmpA’-SpyCatcher	or	Lpp’-OmpA’-
StrepCoreMut2.	However,	upon	induction	and	immunostaining	of	transformed	strains	no	
immunofluorescent	labeling	was	observed,	indicating	that	the	protein	was	not	being	
displayed	on	the	surface	of	the	cells.	We	therefore	decided	to	change	surface	display	
systems	and	obtained	the	plasmid	pAIDA1	through	Addgene.	This	surface	display	system	
utilizes	the	signal	peptide,	and	C-terminal	translocation	unit	of	the	AIDA-I	E.	coli	adhesin	to	
transport	and	display	passenger	domains	on	the	surface	of	E.	coli	(Jarmander	et	al.	2012).	
The	AIDA-I	protein	has	been	used	previously	for	the	expression	of	recombinant	proteins,	
enzymes,	and	enzyme	inhibitors	on	the	cell	surface	(Jose	et	al.	2001,	2002;	Jose	and	Zangen	
2005).	We	designed	and	constructed	the	pAIDA1-SpyCatcher	plasmid	using	by	inserting	
the	 E.	 coli	codon-optimized	SpyCatcher	DNA-fragment	via	Gibson	Assembly	at	the	N-
terminus	of	the	AIDA1c	domain	(Figure	2.2.3A).	Immunostaining	of	E.	coli	transformed	with	
this	construct	with	the	same	combination	of	primary	α-cMyc	antibodies	and	AlexaFluor	
448	conjugated	secondary	antibodies	showed	high	levels	of	fluorescence,	indicating	that	
the	protein	was	being	translocated	to	the	cell	surface	(Figure	2.2.3B).	

	

85	

	

	

86	

Figure	2.2.3:	E.	coli	W	cscR	SpyCatcher	surface	display	construct	

Evaluating	the	ability	of	the	surface	displayed	moieties	with	soluble	conjugate	

(A)	The	pAIDA1-SpyCatcher	construct	was	generated	by	inserting	the	DNA	fragment	
encoding	the	SpyCatcher	into	the	pAIDA1	plasmid	at	the	N-terminus	of	the	AIDA1c	domain.	
(B)	Transforming	this	construct	into	the	E.	coli	W	cscR	strain	and	inducing	the	cells	with	1	
mM	IPTG	triggered	the	production	of	the	fusion	protein.	Cells	were	immunolabeled	with	an	
anti-cMyc	monoclonal	mouse	antibody	and	anti-mouse	IgG	Goat	secondary	antibody	
conjugated	to	AlexaFluor	488.		
proteins	While	the	immunostaining	experiments	indicate	that	the	antigenic	tags	on	both	the	
cyanobacterial	and	heterotrophic	surface	display	proteins	are	accessible,	this	did	not	
directly	confirm	the	binding	activity	of	the	displayed	moieties.	To	directly	assay	the	
capacity	of	surface	tags	to	bind	to	cognate	binding	domains,	we	generated	constructs	
encoding	soluble	recombinant	mNeonGreen	fluorophores	fused	with	either	the	SpyCatcher	
domain	or	a	combination	of	the	SpyTag	and	Strep-tag	II	moieties	(Figure	2.2.4A).		
	

	

Figure	2.2.4:	Protein	reagent	development		

	
(A)	Construct	maps	of	both	pET-11b	plasmids	containing	either	the	mNeonGreen-SpyTag-
StrepTagII	fusion	protein	(Left)	or	the	mNeonGreen-SpyCatcher	fusion	protein	(Right).	(B)	
A	preliminary	EMSA	western	blot	(Left)	and	paired	coomassie	gel	(Right)	that	confirmation	
of	the	purified	mNeonGreen	proteins	are	able	to	interact	through	the	SpyTag	and	
SpyCatcher	domains.	This	membrane	was	incubated	with	anti-cMyc	monoclonal	mouse	
antibody	and	then	the	anti-mouse	IgG	goat	antibody	conjugated	to	Alexafluor	488.	The	
mNG-SpyTag-StrepTagII	runs	at	~35	kDa	on	the	SDS-page	gel	where	as	the	mNG-
SpyCatcher	protein	runs	at	~49	kDa.	In	a	1:1	mixture	of	these	two	protein	reagents	we	see	
the	depletion	of	the	mNG	SpyCatcher	band	and	the	appearance	of	a	band	higher	up	on	the	
gel	at	~74	kDa.	While	slightly	less	than	the	predicted	combined	weight	of	these	two	
proteins	(~84	kDa),	this	shift	indicates	that	these	two	reagents	maintain	affinity	for	one	
another	under	experimental	conditions.	
	
	

87	

	

The	proteins	were	additionally	encoded	with	His6	tags,	expressed	off	of	pET-11b	
plasmids	in	the	 E.	coli	BL21	D3	cell	line.	Thereafter,	the	reporter	proteins	were	purified	
from	cell	lysates	using	a	Nickel-affinity	column.	These	purified	fluorescent	proteins	would	
allow	us	to	examine	whether	the	surface	displayed	SpyTag,	SpyCatcher,	and	StrepCoreMut2	
domains	could	bind	to	freely	diffusible	cognate	binding	partners.	To	verify	activity	of	the	
purified	 mNeonGreen-SpyCatcher	 protein,	 both	
the	 mNeonGreen-SpyCatcher	 and	
mNeonGreen-SpyTag-StrepII	protein	were	mixed	in	equal	ratio.	This	mixture	was	then	
used	to	perform	a	western	blot	mobility	shift	assay	as	the	two	reagents	reacted,	forming	
the	predicted	isopeptide	covalent	bond.	This	assay	demonstrated	that	the	two	recombinant	
proteins	did	indeed	bond	together	(Figure	2.2.4B).		
Given	the	previous	success	of	the	S.	cerevisiae	surface	display	system	(Chapter	2.1)	
and	limited	time	and	resources,	only	the	S.	cerevisiae	strains	were	tested	with	the	purified	
protein	reagent.	The	S.	cerevisiae	SpyCatcher	display	strain	was	induced	in	SGCCA	medium	
(as	in	Chapter	2.1)	and	then	incubated	with	purified	mNeonGreen-SpyTag-StrepII	protein.		
This	demonstrated	that	the	SpyCatcher	domain	is	displayed	in	a	manner	that	is	sufficient	to	
bind	and	react	with	the	diffuse	mNeonGreen-SpyTag-StrepII	protein	(Figure	2.2.5A).	
However,	repeating	the	same	experiment	with	the	 S.	 cerevisiae	StrepCoreMut2	display	
strain	did	not	show	signs	of	mNeonGreen	localization	to	the	cell	surface,	indicating	that	
either	the	StrepCoreMut2	domain	was	non-functional	or	that	the	location	or	orientation	of	
the	Strep-tag	II	on	the	fluorescent	protein	prevented	binding	(Figure	2.2.5B).	Similarly,	our	
initial	attempts	to	localize	the	mNeonGreen-SpyCatcher	protein	to	the	surface	of	the	
cyanobacteria	also	appeared	to	be	unsuccessful	(Figure	2.2.5C).	
	

88	

	

	

Discussion	

Figure	2.2.5:	Preliminary	testing	of	S.	cerevisiae	and	S.	elongatus	strains	with	protein	
reagents	

(A)	Surface	labeling	of	 S.	 cerevisiae	SpyCatcher	with	purified	mNeonGreen-SpyTag-
StrepTagII	protein	demonstrates	that	the	surface	displayed	SpyCatcher	domain	is	
functional	in	this	strain.	(B)	Testing	of	the	 S.	 cerevisiae	StrepCoreMut2	with	the	same	
purified	mNeonGreen-SpyTag-StrepTagII	protein	indicated	that	the	expected	interaction	
was	not	occurring	despite	indications	that	the	recombinant	surface	protein	is	present	on	
the	cell	surface.	(C)	Attempts	at	labeling	of	the	S.	elongatus	SomA-R5SpyTag	strain	with	the	
purified	mNeonGreen-SpyCatcher	protein	also	indicate	that	the	reaction	is	inhibited	by	
either	the	protein	orientation	or	other	steric	hindrances	on	the	cyanobacterial	surface.	
		
In	this	chapter,	I	show	preliminary	data	on	the	diversification	of	peptide	sequences	
utilized	in	the	cyanobacterial	surface	display	system	described	in	Chapter	2.1	and	the	
development	of	heterotrophic	strains	capable	of	displaying	the	corresponding	protein	
	

89	

	

domains	that	can	interact	with	those	peptide	sequences.	Utilizing	the	cyanobacterial	
surface	display	system	I	developed	in	Chapter	2.1,	we	were	able	to	generate	cyanobacterial	
strains	that	translocate	the	engineered	SomA	proteins	containing	both	the	SpyTag	and	
Strep-tag	II	peptide	sequence	to	the	cell	surface	(Figure	2.2.1B).	The	heterotrophic	strains	
of	E.	coli	and	S.	cerevisiae	appear	to	express	their	respective	SpyCatcher	domains	without	
issue	(Figure	2.2.2B).	In	the	case	of	the	S.	cerevisiae	Aga2p-SpyCatcher	display	strain,	the	
SpyCatcher	domain	is	shown	to	be	active	and	able	to	recruit	and	bind	the	SpyTag-
Fluorophore	fusion	protein	purified	in	this	work	(Figure	2.2.4B).	While	surface	display	in	S.	
cerevisiae	has	become	quite	routine	(Pepper	et	al.	2008),	surface	expression	of	the	
SpyCatcher	domain	could	be	a	novel	mechanism	to	recruit	and	permanently	adduct	protein	
based	products	to	the	cell	surface.	
However,	it	appears	in	the	case	of	the	 S.	 elongatus	SomA-R5SpyTag	strain,	the	
accessibility	of	the	SpyTag	sequence	is	restricted	such	that	a	fluorophore	with	the	
SpyCatcher	domain	is	unable	to	interact	with	it	(Figure	2.2.4D).	Surface	layer	stripping	of	
the	cyanobacteria	strains	was	performed	prior	to	removing	extracellular	material	that	
might	directly	occlude	the	surface	of	the	cyanobacteria’s	outer	membrane	(see	Chapter	2.1	
methods).	This	would	suggest	that	the	location	and/or	orientation	of	the	SpyTag	within	the	
extracellular	loop	of	the	recombinant	SomA	may	be	playing	a	role	in	the	inaccessibility	of	
the	sequence.	
The	results	of	Chapter	2.1	and	the	preliminary	data	presented	here	in	Chapter	2.2,	
suggest	that	this	system,	with	further	refinement,	has	the	potential	to	facilitate	interspecies	
adhesion.	We	are	working	toward	interspecies	adhesion	to	artificially	aggregate	different	
species	together	to	examine	whether	this	physical	association	can	stabilize	engineered	
	

90	

	

commensal	or	symbiotic	interactions.	Previous	work	has	shown	that	adhesion	between	
cross-feeding	strains	is	preferred	under	some	conditions	(Marchal	et	al.	2017),	but	not	
others	(Kim	et	al.	2008).	Thus,	having	the	ability	to	induce	tight	and	specific	associations	
between	co-cultured	species	would	provide	a	unique	mechanism	by	which	to	gauge	the	
compatibility	of	two	species.	However,	cell-adhesion	systems	could	be	applied	in	other	
capacities	beyond	co-culture.	A	relevant	example	is	the	sedimentation	of	planktonically	
grown	cyanobacteria.	One	of	the	most	significant	costs	in	microalgal	and	cyanobacterial	
bioindustry	is	accrued	during	the	harvesting	of	biomass	(Sharma	et	al.	2011).	An	active	
area	of	research	is	in	the	use	of	biologically-based	microbial	flocculants	to	sediment	cells	
from	industrial	scale	cultures	(Tong	et	al.	1999).	By	creating	a	specific	adhesion	
mechanism,	it	may	be	possible	to	improve	the	overall	cell	capture	efficiency	of	future	
bioflocculants.		

Materials	and	Methods	

Strains	and	culture	conditions	

Monocultures	of	S.	elongatus	and	S.	cerevisiae	were	cultivated	as	previously	
described	(Chapter	2.1).		E.	coli	strains	were	grown	in	Luria	Bertani	(LB)	medium	with	25	
mg/mL	of	either	Chloramphenicol	(pAIDA1	vector)	or	Carbenicillin	(pET-11b	vector)	to	
maintain	selection	for	the	respective	plasmids,	and	incubated	at	37°C.				

Construct/strain	development	

The	NSIII-SomA-R5Strep-tagII	construct	was	synthesized	using	primers	(F-Strep-
tagII	and	R-Strep-tagII)	that	amplified	a	new	fragment	containing	the	Strep-tagII	sequence	
	

91	

	

that	targeted	the	original	NSII-SomA-R5F	construct	(Chapter	2.1)	as	a	template.	Amplified	
fragments	were	re-circularized	using	Gibson	assembly	described	in	(Chapter	2.1).	The	
NSIII-SomA-R5SpyTag	construct	was	synthesized	by	amplifying	the	NSIII-SomA-R5	
construct	with	R-NS3	and	F-NS3,	and	then	performing	a	Gibson	Assembly	with	at	DNA	
fragment	containing	insert	sequence	(Figure	2.2.1A)	(Source:	IDT).		Transformation	of	WT	
S.	elongatus	PCC	7942	was	performed	as	described	in	(Chapter	2.1.)	Two	iterations	of	both	
SpyCatcher	and	Streptavidin	Core	Mut	2	(E.	coli	or	S.	cerevisiae	codon	optimized)	sequences	
were	synthesized	by	IDT	and	incorporated	into	either	pETCON-pAga2p	or	pAIDA1	via	
Gibson	assembly	(Gibson	et	al.	2009).	The	pETCON-ProteinA	vector	was	cut	with	NdeI	and	
BamHI	to	excise	the	ProteinA	sequence	from	the	plasmid,	the	backbone	was	then	purified	
by	agarose	gel	extraction	(Qiagen).	The	purified	backbone	was	then	used	in	Gibson	
Assembly	reactions	with	either	the	SpyCatcher	or	StrepCoreMut2	gene	fragments	to	create	
the	respective	plasmids.	The	pAIDA1	vector	was	similarly	processed,	cutting	the	vector		
with	the	KpnI	restriction	enzyme,	gel	purifying	the	backbone,	and	assembled	via	Gibson	
reaction	with	the	E.	coli	codon	optimized	SpyCatcher	gene	fragment.	These	vectors	were		

New 

# 

Pseudonym 

Primer 
Direction 

Primer Sequence (5'->3') 

Function 

1 

2 

3 

4 

R-NS3 

Reverse 

TCTAGACATATGTTTTTCCTCCTTGTGTGAAATTGTTATCCGC 

F-NS3 

Forward 

CGGCCGCGTAGGATCCGATCGGCACGTAAGAG 

PCR NSIII Vector with homology 

to somA insert 

PCR NSIII Vector with homology 

to somA insert 

F-Strep-tag 

II 

R-Strep-tag 

II 

Forward 

Reverse 

GGAGCCACCCCCAGTTCGAGAAGGGAGGAGGATCCTCTGAAG
ACACTGGTTCCTTTG 

SomA Lp7 Strep II Tag Insertion 

Forward Primer 

CTCGAACTGGGGGTGGCTCCAGTTGGATCCACCGCCCGAACG
GTTAGAAGCGCTG 

SomA Lp7 Strep II Tag Insertion 

Reverse Primer 

5' Homology Target 

(if applicable) 

somA insert 

somA insert 

somA insert 

somA insert 

Bold = (5' Homology);  Underline= (3' Binding Region) 

Table	2.2.1:	Chapter	2.2	Primer	list	

transformed	into	their	respective	chemically	competent	hosts	(S.	cerevisiae	EBY100	(Lu	
2011)	or	E.	coli	W	cscR	(Inoue	et	al.	1990).		
	

  

92	

Protein	reagent	purification	

Immunostaining	and	western	blotting	

Immunostaining	for	the	S.	elongatus	and	E.	coli	strains	was	performed	as	described	
for	the	cyanobacterial	immunostaining	performed	in	Chapter	2.1.	Immunostaining	of	the	S.	
cerevisiae	strains	were	also	performed	as	described	in	Chapter	2.1.	Western	blotting	was	
also	performed	as	in	Chapter	2.1.	The	relevant	antibodies	used	in	this	chapter	are:	α-FLAG	
tag	monoclonal	mouse	IgG2b	(LT0420;	Lifetein),	α–Strep-tag	II	Mouse/IgG1	
(OAAF04751:Aviva	Systems	Biology),	α-cMyc	Tag	Mouse/IgG2a	(JP_A000704-
100;GenScript),	and	goat	α-mouse	IgG	Alexa	Fluor®	488	(35502;	Thermo).	
Protein	purification	was	performed	using	the	Ni-NTA	Purification	System	
(Novex/LifeTechnologies).	Briefly,	E.	coli	BL21	D3	cells	carrying	either	the	pET-11b	
mNeonGreen-SpyCatcher	or	pET-11b	mNeonGreen-SpyTag-Strep-tagII	plasmids	(Figure	
2.2.4A),	were	inoculated	from	single	colonies	into	125-mL	baffled	flasks	filled	with	50	mL	
of	LB	medium	and	1	mM	IPTG.	These	flasks	were	incubated	~12	hours	in	a	Multitron	Pro	
(Infors)	incubator	at	32°C	with	rotary	shaking	at	150	rpm.	The	cultures	were	then	collected	
in	50	mL	conical	tubes	and	the	cells	were	pelleted	via	centrifugation	in	a	swing-bucket	
centrifuge	for	45	min	at	3,500×g,	while	maintaining	a	temperature	of	4°C.	The	supernatant	
was	removed	and	the	cells	were	resuspended	in	25	mL	of	chilled	PBS	by	vortexing	for	>10	
min.	These	cells	were	then	lysed	utilizing	a	One	Shot	system	cell	disrupter	(Constant	
Systems	Limited)	set	to	a	pressure	level	of	32	kpsi,	and	passed	through	the	sapphire	
aperture	three	times.	The	cell	lysates	were	then	centrifuged	once	more	to	pellet	cellular	
debris	in	an	RC	5C	Plus	ultracentrifuge	(Sorvall)	using	the	SS-34	rotor	(Sorvall).	The	lysates	
	

	

93	

	

were	centrifuged	for	1	hour	at	16,500	rpm	maintained	at	a	temperature	of	4°C.	After	the	
centrifugation,	this	lysate	was	transferred	to	a	pre-chilled	50-mL	conical	tube		
	

	

94	

CHAPTER	3:	BIOTRANSFORMATION	OF	2,4-DINITROTOLUENE	IN	A	PHOTOTROPHIC	

CO-CULTURE	OF	ENGINEERED	SYNECHOCOCCUS	ELONGATUS	AND	PSEUDOMONAS	

PUTIDA	

This	work	has	been	prepared	for	submission	to	Applied	and	Environmental	Microbiology	

Derek T. Fedeson, Pia Saake, Patricia Calero, Pablo Iván Nikel, & Daniel C. Ducat 

Authors: 

	
	
	

	

95	

Author Contributions: 

DTF, PIN, and DCD designed and directed the project and experiments. PC designed and 

engineered the P. putida strains utilized in this work. DTF, PS, and PC performed the 

experiments and recorded the results. DTF, PS, PIN, and DCD analyzed and interpreted the 

data. DTF, PIN, and DCD prepared the manuscript and figures. DTF, PS, PC, PIN, and DCD 

edited the manuscript and figures. 

	

	

	

	

 

Abstract	

In	contrast	to	the	current	paradigm	of	using	microbial	monocultures	in	most	
biotechnological	applications,	increasing	efforts	are	being	directed	towards	engineering	
mixed-species	consortia	to	perform	functions	that	are	difficult	to	program	into	individual	
strains.	Additionally,	the	division	of	labor	between	specialist	species	found	in	natural	
consortia	can	lead	to	increased	catalytic	efficiency	and	stability	relative	to	a	monoculture	or	
a	community	composed	of	generalists.	In	this	work,	we	have	designed	a	synthetic	co-
culture	for	phototrophic	degradation	of	xenobiotics,	composed	of	a	cyanobacterium,	
(Synechococcus	elongatus	PCC	7942)	and	a	heterotrophic	bacterium	(Pseudomonas	putida	
EM173).	Cyanobacteria	fix	CO2	through	photosynthetic	metabolism	and	secrete	sufficient	
carbohydrates	to	support	the	growth	and	active	metabolism	of	P.	putida,	which	has	been	
engineered	to	consume	sucrose	as	the	only	carbon	source	and	to	degrade	the	
environmental	pollutant	2,4-dinitrotoluene	(2,4-DNT).	The	synthetic	consortium	is	able	to	
degrade	2,4-DNT	with	only	light	and	CO2	as	inputs	for	the	system,	and	it	was	stable	over	
time	through	repeated	backdilutions.	Furthermore,	cycling	this	consortium	through	low	
nitrogen	medium	promoted	the	accumulation	of	polyhydroxyalkanoate	(PHA)–an	added-
value	biopolymer–in	P.	putida,	thus	highlighting	the	versatility	of	this	production	platform.	
Altogether,	the	synthetic	consortium	allows	for	simultaneous	bioproduction	of	PHA	and	
remediation	of	the	industrial	pollutant	2,4-DNT,	using	light	and	CO2	as	inputs.	
	

	

96	

	

 

Importance	

In	this	study,	we	have	created	an	artificial	consortium	composed	of	two	bacterial	
species	that	enables	the	degradation	of	the	industrially-produced	environmental	pollutant	
2,4-DNT	while	simultaneously	producing	PHA	bioplastic.	In	these	co-cultures,	the	
photosynthetic	cyanobacteria	fuel	an	engineered	P.	putida	strain	programmed	both	to	use	
sucrose	as	a	carbon	source	and	to	perform	the	biotransformation	of	2,4-DNT.	The	division	
of	labor	in	this	synthetic	co-culture	is	reminiscent	of	that	commonly	observed	in	microbial	
communities	and	represents	a	proof-of-principle	example	of	how	artificial	consortia	can	be	
employed	for	bioremediation	purposes.	Furthermore,	this	co-culture	system	enabled	the	
utilization	of	freshwater	sources	that	could	not	be	utilized	in	classical	agriculture	settings,	
reducing	the	potential	competition	of	this	alternative	method	of	bioproduction	with	
current	agricultural	practices,	as	well	as	remediation	of	contaminated	water	streams.		

	

97	

	

Introduction	

In	nature,	bacteria	typically	co-exist	in	communities	with	hundreds	to	thousands	of	
other	microorganisms,	creating	a	complex	web	of	inter-species	metabolic	reactions	(Little	
et	al.	2008;	Saxena	2015;	Dolinšek	et	al.	2016;	Goldford	et	al.	2017).	Most	consortia	exhibit	
a	high	degree	of	“division	of	labor,”	where	individual	species	have	specialized	metabolisms	
and	exchange	metabolites	and	signals	with	neighbors	(Gestel	et	al.	2015;	Tan	et	al.	2015).	
Interactions	range	from	the	cooperative	degradation	of	toluene	(Nikel	et	al.	2014;	Tecon	
and	Or	2017)	to	the	consumption	of	metabolic	waste	products	(Wilkinson	et	al.	1974).	
Compartmentalization	of	metabolism	across	distinct	species	can	confer	metabolic	
capabilities	on	a	consortium	that	may	be	difficult	to	engineer	within	any	one	individual.	
Additionally,	natural	microbial	consortia	frequently	exhibit	a	high	degree	of	robustness	in	
the	face	of	dynamic	environmental	conditions	and	are	resilient	to	invasive	microbes	
(Kumar	and	Jagadeesh	2016;	Blasche	et	al.	2017).	Thus,	there	has	been	increasing	interest	
in	rationally	engineering	microbial	consortia	for	desired	outputs	by	dividing	metabolic	
pathways	across	species	(Ortiz-Marquez	et	al.	2013).	
Previous	work	from	our	laboratory	and	others	has	focused	on	the	utility	of	strains	of	
cyanobacteria	engineered	to	export	photosynthetically-generated	sucrose	through	the	
heterologous	expression	of	the	cscB	gene	from	Escherichia	coli,	encoding	sucrose	permease	
(Ducat	et	al.	2012;	Du	et	al.	2013;	Song	et	al.	2016).	In	one	study	utilizing	this	engineered	
strain	of	Synechococcus	elongatus	PCC7942	(hereafter,	S.	elongatus	CscB),	sucrose	secretion	
was	such	that	hypothetical	scaled	production	would	significantly	exceed	current	
productivities	of	traditional	sugar	crops	like	sugar	cane,	sugar	beet,	and	corn	(Ducat	et	al.	
	

98	

	

2012).	As	such,	the	strain	has	been	utilized	on	numerous	occasions	as	a	photosynthetic	
module	in	synthetic	co-cultures	as	a	supplier	of	fixed	carbon	for	heterotrophic	partners	
(Smith	and	Francis	2016,	2017;	Hays	et	al.	2017;	Li	et	al.	2017;	Löwe	et	al.	2017a),	
including	a	recent	report	where	co-cultures	were	maintained	for	longer	than	5	months	of	
continuous	co-culture	(Weiss	et	al.	2017).	Communally,	these	works	have	demonstrated	
that	the	 S.	 elongatus	CscB	strain	can	be	flexibly	paired	with	a	variety	of	heterotrophic	
bacteria	[Escherichia	 coli	W	(Hays	et	al.	2017),	 Bacillus	 subtilis	(Hays	et	al.	2017),	
Azotobacter	vinelandii	(Smith	and	Francis	2016,	2017),	Halomonas	boliviensis	(Weiss	et	al.	
2017),	and	Pseudomonas	putida	(Löwe	et	al.	2017a)]	and	yeasts	[Saccharomyces	cerevisiae	
(Ducat	et	al.	2012;	Hays	et	al.	2017),	Cryptococcus	curvatus	(Li	et	al.	2017),	and	Rhodotorula	
glutinis	(Li	et	al.	2017)]	and	demonstrated	that	these	co-cultures	can	be	used	to	
photosynthetically	produce	valued	biological	products,	e.g.,	α-amylase	(Hays	et	al.	2017),	
fatty	 acids	 (Li	 et	 al.	 2017),	 and	 polyhydroxyalkanoates	 (PHA)	
[e.g.,	 poly(3-
hydroxybutyrate)	(PHB)	(Smith	and	Francis	2016,	2017;	Löwe	et	al.	2017a;	Weiss	et	al.	
2017)].	In	work	from	Smith	and	Francis	(Smith	and	Francis	2017)	as	well	as	our	lab	(Weiss	
et	al.	2017)	it	was	shown	that	S.	elongatus	CscB	could	be	immobilized	within	hydrogels;	this	
both	enhanced	carbon	flux	into	sucrose	production	and	allowed	the	cyanobacteria	to	
exchange	diffusible	metabolites	in	medium,	while	enabling	the	heterotrophic	cells	to	be	
readily	harvested	separately.	Taken	together,	this	co-culture	approach	has	potential	as	a	
platform	to	enable	the	modular	photosynthetic	production	of	a	flexible	array	of	
bioproducts.	
A	primary	concern	when	considering	large-scale	aquatic	based	bioproduction,	is	
that	global	potable	water	supplies	are	coming	under	increasing	strain;	scaled	applications	
	

99	

	

of	algae	or	cyanobacteria	would	be	more	sustainable	if	they	can	utilize	marginal	waters	not	
suitable	for	other	agricultural	purposes	(Barry	et	al.	2016).	Industrial	wastewater	streams	
are	one	such	example,	as	they	are	often	contaminated	with	chemical	compounds	toxic	to	
both	flora	and	fauna.	2,4-Dinitrotoluene	(2,4-DNT)	is	a	nitroaromatic	compound	and	is	one	
of	a	diverse	array	of	xenobiotics	that	have	been	released	into	the	environment	due	to	
industrial	synthetic	chemistry	and	other	manufacturing	processes	(Ju	et	al.	2010).	While	
2,4-DNT	is	produced	as	a	by-product	during	polyurethane	and	pesticide	synthesis,	one	of	
the	most	significant	sources	of	these	contaminants	is	explosive	manufacturing	(Spain	
1995).	One	2,4,6-trinitrotoluene	(TNT)-manufacturing	plant	can	contaminate	five	hundred	
thousand	gallons	of	water	with	TNT	and	other	nitroaromatics	in	a	single	day	(Yinon	1990),	
and	at	some	munitions	manufacturing	and	processing	sites,	soil	contamination	is	as	high	as	
200	g	of	TNT	per	kilogram	of	soil	(Hooker	and	Skeen	1999).	These	compounds	are	highly	
stable	within	the	environment,	and	remediation	costs	via	incineration	are	estimated	near		
$400	USD	per	cubic	yard	of	soil	(Griest	et	al.	1998).	
Despite	the	fact	that	these	nitroaromatic	compounds	have	only	recently	been	
introduced	into	the	environment,	a	number	of	bacterial	strains	capable	of	mineralizing	
these	unusual	chemical	structures	have	been	isolated	(Spain	1995;	Symons	and	Bruce	
2006;	Ju	et	al.	2010).	The	most	common	mechanism	by	which	2,4-DNT	is	processed	is	
through	one	or	two-electron	reduction	via	non-specific	nitroreductases	(French	et	al.	
2001).	Enzymes	containing	a	redox-active	flavin	or	iron	center	are	especially	prone	to	
perform	 these	 reactions	 (Bryant	 and	 McElroy	 1991).	 Single	 electron	 reductions	
temporarily	reduces	the	nitro	group	but	it	is	then	immediately	re-oxidized	in	the	presence	
of	molecular	oxygen	resulting	in	the	generation	of	a	superoxide,	the	accumulation	of	which	
	

100	

	

can	lead	to	the	ROS	stress	response	(Park	et	al.	2006).	Two	electron	reductions	result	in	the	
production	of	a	toxic	hydroxylamino	derivative	that	can	react	with	DNA	and	cause	
subsequent	mutations	(Bryant	and	McElroy	1991).	The	nitroso-	and	hydroxylamino-
derivatives	are	even	more	toxic	than	their	parent	molecule,	able	to	form	DNA	and	protein	
adducts	that	lead	to	mutagenesis	and	cellular	damage	(Spain	1995;	Padda	et	al.	2003).	
Additionally,	the	derivatives	from	reduced	nitroaromatics	continue	to	persist	in	the	
environment	(Achtnich	et	al.	1999),	further	emphasizing	the	need	for	alternative	methods	
of	degradation	that	allow	for	the	complete	biotransformation	of	these	compounds.	

Figure	 3.1:	 Conceptual	 design	 of	 the	 photosynthetic	 co-culture	 designed	 for	
simultaneous	biodegradation	and	bioproduction	

This	co-culture	of	alginate-encapsulated	 S.	 elongatus	CscB	and	 P.	 putida	EM!DNT!S	
photosynthetically	drives	the	biodegradation	of	the	toxin	2,4-dinitrotoluene	(2,4-DNT)	and	
simultaneously	produce	the	bioplastic	polyhydroxyalkanoate	(PHA).	The	S.	elongatus	CscB	
embedded	within	alginate	beads,	perform	photosynthesis,	fixing	carbon	dioxide	from	the	
air	using	photosynthetically	active	radiation	(PAR).	This	fixed	carbon	is	converted	into	
sucrose	that	is	then	exported	from	the	cells	in	to	the	culture.	There	the	sucrose	is	
consumed	by	the	P.	putida	EM!DNT!S	which	uses	this	energy	to	biodegrade	the	toxin	2,4-
DNT.	Additionally,	the	P.	putida	EM!DNT!S	can	also	be	directed	to	accumulate	PHA	in	their	
biomass,	making	this	system	multifunctional.	
	
101	

	

	

A	separate,	oxidative	pathway	for	degrading	2,4-DNT	was	identified	in	Burkholderia	
sp.	R34,	a	strain	isolated	from	surface	water	contaminated	by	an	ammunition	waste	plant	
(Spanggord	et	al.	1991).	The	gene	cluster	responsible	for	processing	2,4-DNT	has	been	
identified	(Spanggord	et	al.	1991;	Haigler	et	al.	1994;	Nishino	and	Paoli	2000)	and	contains	
7	genes.	Enzymes	within	the	 dnt	degradation	pathway	are	homologs	of	enzymes	that	
function	as	part	of	an	existing	naphthalene	degradation	pathway	in	Burkholderia	sp.	R34	
and	proceed	through	oxidative	steps	that	release	nitrite.	This	suite	of	genes	has	since	been	
chromosomally	integrated	into	a	strain	of	Pseudomonas	putida	EM173	via	a	Tn7	construct	
to	facilitate	the	analysis	of	this	pathway	(Pérez-Pantoja	et	al.	2013;	Akkaya	et	al.	2018).	P.	
putida	is	generally	considered	a	reliable	chassis	for	studying	the	biodegradation	of	organic	
compounds	due	to	its	tolerance	of	organic	solvents	(Ramos	et	al.	2002)	and	versatile	
central	metabolism	(Nelson	et	al.	2002;	Nikel	et	al.	2016).	
In	this	study,	we	explore	whether	the	aforementioned	synthetic	consortium	method	
(Weiss	et	al.	2017)	can	be	engineered	to	utilize	and	remediate	water	streams	contaminated	
with	 the	 environmental	 pollutant	 2,4-DNT	 while	 also	 producing	 the	 bioplastic	
polyhydroxyalkanoate	(PHA)	(Figure	3.1).	This	was	accomplished	through	the	pairing	of	an	
engineered	strain	of	P.	putida,	containing	the	genes	needed	to	both	metabolize	sucrose	and	
to	degrade	2,4-DNT	with	the	sucrose-exporting	S.	elongatus	CscB	encapsulated	in	alginate	
hydrogel	beads.	Approaching	the	bioremediation	of	this	compound	via	the	synthetic	
consortium	method	allows	for	the	bioprocess	to	be	photosynthetically	powered	while	
avoiding	the	complexities	of	introducing	a	new	enzymatic	pathway	into	photosynthetic	
cyanobacteria.	We	demonstrated	that	these	co-cultures	can	successfully	execute	the	
biotransformation	of	2,4-DNT	via	the	engineered	pathway	and	are	also	able	to	accumulate	
	

102	

	

PHA	as	a	secondary	function.	This	work	is	proof	of	principle	for	the	use	of	synthetic	
cyanobacteria/heterotroph	consortia	for	combined	bioremediation	and	bioproduction	
applications.	

Results	

Alginate-encapsulated	S.	elongatus	CscB	can	tolerate	2,4-DNT	at	higher	

concentrations	than	planktonic	cultures		

As	2,4-DNT	is	known	to	be	highly	toxic	to	a	range	of	biological	organisms	(Yoon	et	
al.	2006;	Rocheleau	et	al.	2010),	we	first	characterized	the	effect	of	this	industrial	
byproduct	on	the	growth	and	viability	of	cultures	of	 S.	 elongatus	CscB.	We	measured	
culture	density	and	the	level	of	chlorophyll	 a	(Chl	 a)	in	planktonic	 S.	 elongatus	CscB	
cultures	in	the	presence	of	increasing	concentrations	of	2,4-DNT	(Figure	3.2AB).	Even	the	
lowest	concentration	of	2,4-DNT	examined	(8	µM)	caused	a	~60%	growth	impairment	in	
planktonic	cyanobacterial	cultures	(Figure	3.2A),	and	a	near	cessation	of	growth	in	the	first	
48	h	was	observed	at	2,4-DNT	concentrations	≥	15	µM	(Figure	3.2A).	The	toxicity	of	this	
compound	at	these	concentrations	is	highly	relevant	as	leachates	from	soil	contaminated	
with	TNT	and	its	breakdown	products	have	been	recorded	as	high	as	98	µM	(Griest	et	al.	
1995).	Chl	 a	concentration	is	commonly	measured	as	a	proxy	for	physiological	stress	in	
cyanobacteria	and	is	influenced	by	a	variety	of	conditions	(Sauer	et	al.	2001;	Latifi	et	al.	
2009;	Korosh	et	al.	2018).	Cultures	at	concentrations	of	2,4-DNT	ranging	from	31	μM	to	
125	μM	exhibited	an	overall	loss	of	Chl	 a	(Figure	3.2B).	This	progressive	loss	of	Chl	 a	
aligned	with	visual	chlorosis	and	bleaching	of	these	cultures;	this	observation	was	
consistent	at	multiple	2,4-DNT	concentrations	and	at	higher	starting	culture	densities	
	

103	

	

(Figure	3.S1	in	the	Supplemental	Materials).	These	preliminary	experiments	indicated	that	
planktonic	cyanobacteria	are	highly	sensitive	to	even	low	concentrations	of	2,4-DNT,	which	
would	complicate	their	ability	to	be	engineered	to	directly	degrade	this	nitroaromatic.	
Further,	the	viability	loss	of	planktonic	S.	elongatus	CscB	would	need	to	be	mitigated	for	
any	co-culture	applications	targeting	the	remediation	of	these	compounds.	

Figure	 3.2:	 Growth	 and	 physiological	 parameters	 of	 planktonic	 and	 alginate-
encapsulated	S.	elongatus	CscB	

	
(A)	Planktonic	cultures	of	S.	elongatus	CscB	inoculated	at	OD750	=	0.2	were	added	with	2,4-
DNT	at	0	μM-125	μM	and	incubated	for	54	h.	Bacterial	growth	was	estimated	from	OD750	
readings.	(B)	Chlorophyll	a	content	of	cultures	normalized	to	the	values	in	control	cultures	
with	no	added	2,4-DNT	at	each	time	point.	(C)	Chlorophyll	 a	content	of	encapsulated	 S.	
elongatus	CscB	cells	normalized	to	the	values	in	uninduced	(–),	control	cultures	with	no	
added	2,4-DNT.	(D)	Total	sucrose	concentration	in	the	culture	supernatant	of	induced	(1	
mM	IPTG)	alginate-encapsulated	S.	elongatus	CscB	while	grown	with	2,4-DNT	at	either	0	
μm,	125	μM,	or	250	μM.	For	the	experiments	shown	in	(A-C),	the	mean	values	for	n	=	3	are	
indicated,	and	error	bars	represent	standard	deviations;	for	the	experiment	indicated	in		
	
	

		
104	

	

Figure	3.2	(cont’d)	

(D),	the	mean	values	for	n	=	3	with	3	technical	replicates	per	condition	are	indicated,	and	
error	bars	represent	standard	deviations.	
		
	In	previous	work	(Weiss	et	al.	2017),	we	utilized	alginate	hydrogel	encapsulation	of	
S.	 elongatus	CscB	to	stabilize	a	co-culture	with	 Halomonas	 boliviensis	under	prolonged	
nitrogen	stress	conditions.	Encapsulation	has	been	used	for	the	immobilization	of	a	variety	
of	cell	types,	and	has	often	led	to	increased	stress	tolerance,	cell	longevity,	and	metabolic	
flux	toward	target	end	products	(Bailliez	et	al.	1985;	Gillet	et	al.	2000;	Srinivasulu	et	al.	
2003;	Leino	et	al.	2012;	Therien	et	al.	2014;	Ruiz-güereca	and	Sánchez-saavedra	2016).	
Alginate-encapsulated	 S.	elongatus	CscB	cells	did	not	exhibit	chlorosis	in	the	presence	of	
2,4-DNT,	even	when	the	concentration	was	raised	to	250	μM,	near	the	solubility	limit	for	
this	compound	(Figure	3.S2	in	the	Supplemental	Materials).	We	then	exposed	encapsulated	
S.	 elongatus	CscB	cells	to	2,4-DNT	at	250	μM	for	7	days	while	simultaneously	inducing	
expression	of	the	CscB	exporter.	Chl	 a	was	extracted	from	the	beads	and	measured	via	
spectrophotometry	(Figure	3.2C).	The	data	show	that	while	the	induction	of	 cscB	
expression	leads	to	a	slight	decrease	in	relative	Chl	 a	levels,	the	addition	of	2,4-DNT	to	
alginate-encapsulated	cyanobacteria	did	not	further	decrease	Chl	 a	concentration.	
Similarly,	the	Chl	 a	concentration	per	cell	was	maintained	at	a	level	similar	to	that	of	
planktonic	cells	grown	under	our	laboratory	conditions	(Figure	3.S3	in	the	Supplemental	
Materials).	Finally,	we	measured	sucrose	export	rates	from	IPTG-induced,	encapsulated	S.	
elongatus	CscB	in	the	presence	of	increasing	2,4-DNT	concentrations	(Figure	3.2D).	Sucrose	
export	was	maintained	for	multiple	days	despite	exposure	to	125	or	250	µM	2,4-DNT.	
Altogether,	alginate	encapsulation	appears	to	stabilize	S.	elongatus	when	exposed	to	high	
levels	of	2,4-DNT	over	long	time	periods.	
	

105	

 

	

106	

Engineering	P.	putida	EM173	for	sucrose	consumption	and	evaluation	of	growth	

parameters	in	the	presence	of	alginate-encapsulated	S.	elongatus	CscB	

We	next	set	to	construct	strains	of	 P.	 putida	that	can	utilize	sucrose	as	the	only	
carbon	source	and	are	capable	of	degrading	2,4-DNT.	P.	putida	does	not	normally	utilize	
sucrose	as	a	carbon	substrate.	The	specific	strains	we	used	are	derivatives	of	the	
genetically-tractable,	prophage-less	P.	putida	strain	EM173	(Martínez-García	et	al.	2015).	
To	enable	sucrose	consumption	by	P.	putida	EM173,	we	first	transformed	this	strain	with	
plasmid	pSEVA221-cscRABY	(Löwe	et	al.	2017b,	2018),	bearing	the	sucrose	utilization	
genes	from	 P.	 protegens	Pf-5	(Figure	3.3A).	Specifically,	this	plasmid	constitutively	
expresses	genes	encoding	a	sucrose	hydrolase	(CscA,	PFL_3237)	and	a	sucrose	permease	
(CscB,	PFL_3238),	along	with	a	cognate	transcriptional	regulator	(CscR,	PFL_3236).	The	
introduction	of	pSEVA221-cscRABY	into	strain	EM173	(giving	rise	to	 P.	 putida	EM·S)	
enabled	catabolism	of	sucrose	and	bacterial	growth	from	the	disaccharide.	Separately,	a	
synthetic	mini-Tn7	transposon,	carrying	the	functions	required	for	2,4-DNT	degradation	in	
Burkholderia	sp.	R34	(Figure	3.3A)	was	constructed	as	described	elsewhere	(Akkaya	et	al.	
2018).	The	dnt	gene	cluster	in	this	transposon	was	delivered	into	the	unique	att·Tn7	site	
within	the	chromosome	of	P.	putida	EM·S,	resulting	in	a	stable,	engineered	strain	designed	
for	2,4-DNT	degradation	and	sucrose	consumption	(P.	 putida	EM·DNT·S).	To	test	for	
successful	catabolism	of	sucrose	by	these	strains,	we	performed	an	initial	characterization	
of	sucrose	catabolism	in	an	experiment	in	which	M9	minimal	medium	with	20	g/L	sucrose	
was	inoculated	with	either	 P.	 putida	EM·S	or	 P.	 putida	EM·DNT·S	at	an	optical	density	
measured	at	600	nm	(OD600)	=	0.1.	These	cultures	were	followed	over	the	course	of	24	h,	
monitoring	both	culture	density	(OD600)	and	soluble	sucrose	concentrations	(Figure	3.3B).	
	

	

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
	
	

Figure	 3.3:	 Construction	 of	 P.	 putida	 EM·DNT·S	 and	 characterization	 of	 sucrose-
dependent	growth	alone	or	in	co-cultures	

(A)	The	2,4-DNT	degradation	gene	cluster	from	 Burkholderia	sp.	R34.	(B)	Growth	of	 P.	
putida	EM·DNT·S	at	different	concentrations	of	sucrose	(0,	1.25,	2.5,	5,	and	10	g/L)	in	M3	
medium.	(C)	P.	putida	strains	EM·S	and	P.	putida	EM·DNT·S	were	grown	in	M3	media	with	
20	g/L	sucrose	overnight	and	then	inoculated	at	OD600	~	0.1	into	flasks	containing	alginate	
beads	with	or	without	encapsulated	S.	elongatus	CscB,	all	cultures	contained	1	mM	IPTG	for	
induction	of	sucrose	export.	OD600	measurements	were	taken	over	the	course	of	216	hours,	
tracking	the	growth	of	the	P.	putida	strains	in	the	co-culture.	At	96	h	post-inoculation,	all	of	
the	M3	media	was	removed	and	replaced	with	fresh	medium,	allowing	the	residual	 P.	
putida	cells	on	the	surface	of	the	alginate	beads	to	repopulate	the	culture.	For	the		
	

107	

	

Figure	3.3	(cont’d)	

experiments	shown	in	(B)	and	(C),	the	mean	values	for	n	=	3	are	indicated,	and	error	bars	
represent	standard	deviations.	
		Both	P.	putida	EM·S	or	P.	putida	EM·DNT·S	grew	exponentially	over	the	course	of	the	first	9	
h	post-inoculation	and	by	24	h	had	reached	final	densities	of	OD600	=	4.3	and	6	respectively.	
During	this	time,	sucrose	concentration	had	declined	in	a	near	linear	fashion	from	20	g/L	to	
<	3	g/L	at	24	h.	As	sucrose	was	the	only	added	carbon	source	and	we	could	clearly	observe	
the	consumption	of	sucrose	over	time,	these	data	definitively	demonstrate	that	the	
pSEVA221-cscRABY	vector	enabled	sucrose	utilization	in	these	two	strains.	
We	cultured	our	doubly-modified	P.	putida	EM·DNT·S	in	the	presence	of	a	range	of	
sucrose	concentrations	to	determine	if	this	strain	is	capable	of	growing	on	a	minimal	
medium	designed	for	cyanobacterial	growth	with	sucrose	as	a	sole	carbon	source.	For	this	
purpose,	we	generated	a	phosphate	buffered	minimal	medium	derived	from	BG-11,	herein	
referred	to	as	M3	medium	(Table	S1	in	the	Supplemental	Material).	To	gauge	the	growth	
capacity	of	P.	putida	EM·DNT·S	under	these	conditions,	we	inoculated	M3	medium	that	did	
not	contain	a	carbon	source,	and	incubated	cells	overnight.	This	promoted	acclimation	to	
the	medium	and	depletion	of	internal	carbon	storage	compounds	that	could	confound	the	
accurate	determination	of	growth	in	the	M3	medium.	These	cells	were	washed	with	fresh	
M3	medium	before	being	inoculated	into	culture	flasks	with	a	range	of	sucrose	
concentrations	(0	g/L	to	10	g/L)	and	growth	was	tracked	for	54	h	(Figure	3.3C).	Bacterial	
growth	was	evident	at	sucrose	concentrations	ranging	from	1.25-10	g/L	(Figure	3.3C),	
though	carbon	may	have	been	growth-limiting	at	concentrations	lower	than	2.5	g/L	(Figure	
3.3C).	These	results	confirm	that	the	heterologous	expression	of	the	cscRABY	genes	from	P.	
protegens	Pf-5	is	sufficient	to	confer	sucrose	utilization	on	 P.	 putida	in	our	background	
	

108	

	

strain	bearing	the	2,4-DNT	degradation	gene	cluster,	and	indicated	that	the	engineered	P.	
putida	strain	can	grow	in	the	cyanobacterial	M3	medium―	setting	the	basis	for	conducting	
co-cultures.	

Growth	of	engineered	P.	putida	strains	is	supported	by	sucrose-rich	cyanobacterial	

exudates	in	a	synthetic	consortium	system	

We	next	explored	how	the	engineered	P.	putida	strains	(P.	putida	EM·S	and	P.	putida	
EM·DNT·S)	behaved	in	co-culture.	The	 P.	putida	strains,	grown	overnight	in	M3	medium	
with	20	g/L	sucrose,	were	inoculated	at	OD600	~	0.1	into	culture	flasks	containing	either	
empty	alginate	beads	or	alginate	beads	with	encapsulated	S.	elongatus	CscB	(Figure	3D).	In	
the	first	48	h,	both	P.	putida	strains	the	OD600	continued	to	increase	in	all	flasks,	despite	the	
absence	of	a	carbon	source	in	the	flask	containing	empty	alginate	beads.	Internal	stores	of	
carbon	likely	drove	this	residual	growth	in	 P.	 putida.	However,	these	carbon	stores	
appeared	to	be	depleted	after	this	period	of	time	as	the	optical	density	of	the	cultures	
containing	empty	alginate	beads	declined	over	the	next	24	h	(Figure	3.3D).	At	96	h,	the	
culture	supernatant	was	removed	from	all	co-cultures,	leaving	the	 S.	 elongatus	CscB	
alginate	beads	in	place.	Fresh	medium	was	added	and	the	co-culture	densities	were	tracked	
for	another	120	h	(Figure	3.3D).	Only	the	 P.	putida	strains	in	flasks	with	alginate	beads	
containing	 S.	elongatus	CscB	showed	signs	of	regrowth	after	the	medium	exchange,	with	
OD600	increasing	from	0.1	at	96	h	to	0.5-0.6	at	216	h.	These	results	demonstrate	the	 P.	
putida	strains	tolerate	and	can	utilize	the	exudates	from	the	beaded	S.	elongatus	CscBs,	and	
are	not	significantly	utilizing	the	alginate	beads	as	a	carbon	source.	These	results	are	in	
agreement	with	previous	work	that	utilized	P.	putida	EM·S	and	S.	elongatus	CscB	(Löwe	et	
	

109	

	

al.	2017a),	indicating	that	these	species	are	compatible	under	light-driven	co-culture	
conditions.		

Biotransformation	of	2,4-DNT	by	engineered	P.	putida	EM·DNT·S	in	both	

monoculture	and	co-culture	

We	next	examined	the	functionality	of	the	2,4-DNT	degrading	gene	cluster	in	 P.	
putida	monocultures	in	M3	medium	supplemented	with	2	g/L	sucrose	as	the	sole	carbon	
source.	The	dnt	cluster	is	of	significant	interest	to	numerous	research	groups	due	to	the	fact	
that	it	is	an	actively	evolving	pathway	(de	las	Heras	et	al.	2011;	Nikel	and	Chavarrı	2013)	
and	its	unique	biological	processing	of	2,4-DNT	(Symons	and	Bruce	2006).	This	enzymatic	
pathway	enables	the	oxidative	degradation	of	2,4-DNT,	relying	on	two	key	successive	
dioxygenations	of	the	2,4-DNT	substrate	mediated	by	DntA	and	DntB	(Figure	3.4A).	2,4-
DNT	is	first	dioxygenated	by	DntA	to	form	4-methyl-5-nitrocatechol	(4M5NC),	a	compound	
with	a	strong	absorption	peak	at	420	nm	(de	las	Heras	et	al.	2011).	4M5NC	is	the	substrate	
of	DntB,	which	catalyzes	another	oxygenation	reaction,	transforming	4M5NC	into	2-
hydroxy-5-methylquinone	(2H5MQ),	an	intermediate	with	an	absorption	peak	at	485	nm	
(de	las	Heras	et	al.	2011).	2H5MQ	is	then	processed	by	a	number	of	additional	enzymes	
encoded	by	genes	in	the	pathway	allowing	for	the	complete	mineralization	of	2,4-DNT	
(Figure	3.4A)	(Spanggord	et	al.	1991;	Nishino	and	Paoli	2000).		
We	examined	the	transformation	of	2,4-DNT	by	either	P.	putida	EM·S	or	P.	putida	
EM·DNT·S	strains	over	time	(Figure	3.4B-E).	Flasks	containing	P.	putida	EM·S	or	P.	putida	
EM·DNT·S	were	inoculated	at	OD600	~	0.1	and	2,4-DNT	was	added	at	250	μM	to	the	
cultures.	The	supernatant	of	the	P.	putida	EM·DNT·S	cultures	was	visibly	changed	during		
	

110	

	

	

111	

Figure	3.4:	2,4-DNT	biotransformation	in	monocultures	of	engineered	P.	putida	

	

Figure	3.4	(cont’d)	

Figure	3.4:	2,4-DNT	biotransformation	in	monocultures	of	engineered	P.	putida	

(A)	Exogenous	pathway	for	the	oxidative	degradation	of	2,4-DNT.	(B)	(Left)	Representative	
flask	cultures	of	monocultures	of	P.	putida	EM!DNT!S	and	P.	putida	EM!S	over	22	h	in	M3	
medium	with	20	g/L	sucrose	with	the	addition	of	250	μM	2,4-DNT.	(Right)	Enhanced	side-
by-side	comparison	of	the	cultures.	(E)	The	select	time	points	displayed	of	average	culture	
supernatant	 spectrum	 (n	 =	 3)	 of	 P.	 putida	 EM!DNT!S	 measured	 via	 scanning	
spectrophotometry.	(D)	Representative	LC/MS	elution	profile	of	P.	putida	EM!DNT!S	and	P.	
putida	EM!S	supernatant	after	4	hours	of	incubation	M3	medium	with	2	g/L	sucrose	and	
250	μM	2,4-DNT.	4M5NC	elutes	from	the	column	after	2.3	min	negative	ion	mode.	(E)	
Quantification	of	4M5NC	in	supernatants	of	both	P.	putida	EM!S	and	P.	putida	EM!DNT!S	
monocultures	at	the	4	h	via	LC/MS.	P.	putida	EM!S	strain	did	not	generate	any	detectable	
amount	of	4M5NC,	while	the	 P.	 putida	EM!DNT!S	strain	accumulated	4M5NC.	(F)	
Disappearance	of	2,4-DNT	in	 P.	putida	monocultures	grown	in	M3	medium	with	20	g/L	
sucrose	and	250	μM	2,4-DNT	measured	by	GC/MS	(n	=	3,	reps	error	bars	represent	
standard	deviations).	
		the	experiment	(Figure	3.4B),	and	these	spectroscopic	shifts	were	consistent	with	the	
accumulation	of	intermediates	of	2,4-DNT	breakdown	through	the	exogenous	oxidative	
pathway	(Figure	3.4A).	As	soon	as	2	h	after	addition	of	2,4-DNT	to	the	culture	medium	at	
250	μM,	the	supernatant	turned	yellow	(Figure	3.4B),	which	is	consistent	with	the	
accumulation	of	the	first	pathway	intermediate,	4M5NC	(de	las	Heras	et	al.	2011).	Four	
hours	
later,	the	supernatant	became	visibly	orange	(Figure	3.4B),	suggestive	of	
accumulation	of	the	second	intermediate,	2H5MQ	(de	las	Heras	et	al.	2011).	These	changes	
in	supernatant	coloration	were	not	observed	in	the	control	reactions	with	the	non-
degrading	P.	putida	EM·S	strain	(Figure	3.4B).	
We	further	verified	that	the	colorimetric	changes	in	the	supernatant	of	 P.	 putida	
EM·DNT·S	exposed	to	2,4-DNT	could	be	attributed	to	breakdown	of	the	compound	through	
the	heterologous	 dnt	pathway.	An	8h	time-course	experiment	was	performed	with	
additional	time	points	in	which	both	strains	of	P.	putida	were	grown	in	M3	medium	with	2	
g/L	sucrose	during	which	the	cultures	were	sampled	and	the	supernatant	extracted.	The	
	

112	

	

absorbance	spectra	of	P.	putida	EM·DNT·S	supernatants	exhibited	a	characteristic	rise	in	an	
absorption	peak	at	420	nm	over	time	until	4	h,	at	which	time	a	second	absorption	peak	at	
485	nm	became	evident	(Figure	3.4C).	No	defined	peaks	were	evident	in	the	visible	
wavelength	absorption	spectra	of	the	non-degrading	control	strain	(P.	putida	EM·S)	(Figure	
3.S4	in	the	Supplemental	Materials).	Although	spectroscopic	analysis	is	well-supported	in	
the	literature	as	a	metric	for	measuring	activity	of	this	oxidative	pathway	(de	las	Heras	et	
al.	2011),	a	more	direct	method	of	quantification	was	desirable	to	confirm	the	appearance	
of	the	pathway	intermediates.	Hence,	supernatant	samples	from	both	P.	putida	EM·S	and	
EM·DNT·S	cultures	at	the	4h	time	point	were	evaluated	for	the	presence	of	4M5NC	by	
liquid	chromatography	coupled	to	mass	spectrometry	(LC/MS).	Figure	3D	shows	a	
representative	LC/MS	profile	from	one	of	the	 P.	 putida	EM·DNT·S	cultures,	the	peak	
representing	4M5NC	is	indicated.	This	peak	matches	that	of	the	4M5NC	standard	included	
for	quantification	(Figure	3.S5	in	the	Supplemental	materials),	and	comparison	to	cultures	
of	the	non-degrading	strain	indicate	that	there	was	no	4M5NC	generated	by	the	control	
strain	(Figure	3.4D-E).	
To	confirm	specific	degradation	of	the	2,4-DNT,	culture	supernatants	were	
periodically	sampled	over	two	days.	2,4-DNT	was	observed	to	be	rapidly	lost	from	P.	putida	
EM·DNT·S	cultures	over	the	course	of	the	first	4	h,	as	measured	by	tandem	gas	
chromatography	coupled	to	mass	spectrometry	(GC-MS),	and	had	dropped	to	undetectable	
levels	by	22	h,	a	kinetic	pattern	of	2,4-DNT	transformation	similar	to	what	has	been	
reported	for	 Burkholderia	sp.	R34	(Pérez-Pantoja	et	al.	2013)	and	other	engineered	 P.	
putida	strains	(Akkaya	et	al.	2018)	(Figure	3.S6A	in	the	Supplemental	Materials).	2,4-DNT	
concentrations	in	cultures	containing	the	P.	putida	EM·S	strain	appeared	to	decrease	at	a		
	

113	

	

 

 
Figure	3.5:	Degradation	of	2,4-DNT	in	co-culture	and	long-term	co-culture	cycling	

(A)	Growth	of	P.	putida	strains	in	co-culture	with	alginate	bead	encapsulated	S.	elongatus	
CscB	with	250	μM	2,4-DNT	present	in	the	media.	(B)	GC/MS	analysis	of	the	supernatants	
from	the	24	hr	co-culture	demonstrates	that	presence	of	living	cells	triggers	the	
transformation	of	2,4-DNT	via	either	the	non-specific	reductive	pathway	or	exogenous	
oxidative	pathway	such	that	the	concentration	of	2,4-DNT	has	dropped	to	undetectable	
levels	by	24	hours	post	inoculation.	(C)	In	a	separate	experiment,	we	performed	LC/MS	
quantification	of	4M5NC	in	beaded	co-cultures.	Within	1	h	of	adding	2,4-DNT	into	beaded	
co-cultures,	P.	putida	EM!DNT!S	cultures	accumulated	4M5NC	that	is	been	fully	degraded	
by	24	h.	(D)	Optical	density	of	a	long-term	co-culture	of	P.	putida	strains	with	encapsulated	
S.	elongatus	CscB	with	250	μM	2,4-DNT.	This	displayed	segment	of	this	two	week	long	co-
culture	shows	stable	cycling	of	the	 P.	putida	every	4-5	days.	At	each	cycling,	the	culture	
supernatant	containing	the	 P.	 putida	is	removed	and	fresh	media	is	added	back	to	the	
culture	along	with	1	mM	of	IPTG,	to	maintain	induction	of	cyanobacteria	sucrose	export,	
and	250	μM	2,4-DNT.	For	(A-C),	n	=	3,	error	bars	represent	standard	deviations.	
	similar	rate	(Figure	3.S6A	in	the	Supplemental	Materials).	This	loss	of	2,4-DNT	may	be	
attributed	to	the	adsorption	of	the	2,4-DNT	to	cell	surfaces,	a	known	property	of	
nitroaromatic	compounds	that	makes	them	difficult	to	extract	from	biological	substrate	
	

114	

	

of	

for	

the	

presence	

(Achtnich	 et	 al.	 1999),	 and/or	 the	 reduction	 of	 the	 compound	 by	 non-specific	
nitroreductases.	Previous	work	by	Akkaya	et	al.	(Akkaya	et	al.	2018)	with	the	P.	putida	EM	
strain	indicated	that	the	reductive	pathway	(Figure	3.S6C	in	the	Supplemental	Materials)	is	
not	likely	a	major	contributor	to	2,4-DNT	transformation.	We	assessed	our	supernatant	
2-amino-4-nitrotoluene/4-amino-2-nitrotoluene	
samples	
(2A4NT/4A2NT),	intermediates	in	the	non-specific	reductive	pathway	(Figure	3.S6D	in	the	
Supplemental	Materials).	These	compounds	were	detected	in	minor	amounts	(<	3	μM)	in	
both	the	P.	putida	EM·DNT·S	and	P.	putida	EM·S	cultures	(Figure	3.S6D	in	the	Supplemental	
Materials).	While	it	is	difficult	to	quantify	the	flux	attributable	to	the	reductive	pathway,	the	
accumulation	of	2A4NT/4A2NT	compounds	was	reduced	by	22%	in	 P.	putida	EM·DNT·S	
cultures	compared	to	the	P.	putida	EM·S	(Figure	3.S6D	in	the	Supplemental	Materials). 

 

Degradation	of	2,4-DNT	by	a	synthetic	consortium	of	P.	putida	EM·DNT·S	and	

alginate-encapsulated	S.	elongatus	CscB	and	long-term	culture	potential	

We	introduced	2,4-DNT	into	co-cultures	of	P.	putida	and	encapsulated	S.	elongatus	
CscB.	Both	the	P.	putida	EM·S	and	P.	putida	EM·DNT·S	strains	were	able	to	grow	with	the	
beaded	 S.	 elongatus	CscB	in	the	presence	of	250	μM	2,4-DNT	(Figure	3.5A).	Cultures	
containing	only	encapsulated	S.	elongatus	CscB	or	empty	alginate	beads	served	as	controls	
for	both	optical	density	measurements	as	well	as	any	potential	2,4-DNT	adsorption.	GC-MS	
analysis	for	2,4-DNT	content	of	these	cultures	demonstrated	that	cultures	containing	 P.	
putida	or	 S.	 elongatus	cells	removed	2,4-DNT	from	the	culture	(Figure	5B),	in	line	with	
previous	results.	In	contrast,	2,4-DNT	concentrations	did	not	decrease	in	the	flasks	
containing	only	empty	beads	(Figure	3.5B).	Co-cultures	containing	 P.	 putida	EM·DNT·S	
	

115	

	

demonstrated	a	similar	color	change	in	the	culture	supernatant	to	that	of	 P.	 putida	
EM·DNT·S	monocultures,	indicating	that	the	exogenous	pathway	retained	its	function	in	the	
consortium	system.	These	results	demonstrate	that	this	system	can	provide	a	directed	
method	for	the	photosynthetically-driven	degradation	of	2,4-DNT.	Subsequent	LC-MS	
testing	of	co-culture	supernatants	shortly	(1	h)	and	24	h	after	inoculation	of	the	co-culture	
with	2,4-DNT	revealed	that,	as	in	previous	monoculture	experiments,	the	 P.	 putida	
EM·DNT·S	co-cultures	accumulated	4M5NC	(Figure	3.5C).	This	observation,	again,	indicates	
that	2,4-DNT	is	being	degraded	via	the	oxidative	pathways	in	these	co-cultures.	
	
We	initiated	a	longer	duration	co-culture	in	which	we	inoculated	both	strains	of	P.	
putida	into	culture	flasks	with	encapsulated	 S.	 elongatus	CscB	cells	and	followed	the	
cultures	for	15	days,	exchanging	the	culture	supernatant	every	4-5	days	with	fresh	medium	
containing	250	μM	2,4-DNT	(Figure	3.5D).	Regrowth	of	 P.	 putida	strains	following	
backdilution	indicate	stable	repopulation,	allowing	for	continual	degradation	of	the	2,4-
DNT	(Figure	3.5D).	The	total	mass	of	2,4-DNT	cleared	by	the	 P.	putida	EM·DNT·S	strain	
over	this	period	amounts	to	ca.	4	mg.	

in	a	synthetic	consortium.	

 

Simultaneous	2,4-DNT	biodegradation	and	PHA	bioproduction	by	engineered	strains	

P.	putida	accumulates	PHA	as	a	carbon	storage,	particularly	under	nitrogen-depleted	
conditions	(Löwe	et	al.	2017a),	providing	the	opportunity	to	both	bioremediate	2,4-DNT	
and	simultaneously	produce	a	valuable	byproduct.	P.	putida	cells	were	inoculated	into	the	
encapsulated	and	induced	S.	elongatus	CscB	culture	flasks	at	OD600	~	0.5	with	250	μM	2,4-
DNT	(Figure	3.6A).	These	flasks	contained	either	M3	or	M3-N	medium,	the	latter	of	which		
	

116	

	

 
 
 
 
 
 
 
 
 

Figure	3.6:	PHA	accumulation	in	P.	putida	bioremediating	co-cultures	

(A)	Co-cultures	of	beaded	S.	elongatus	CscB	and	P.	putida	strains	in	the	presence	of	2,4-DNT	
with	the	standard	or	reduced	concentration	(2	mM)	of	nitrogen.	Nitrogen	scarcity	triggers	
P.	putida	to	allocate	excess	carbon	into	polyhydroxyalkanoate	(PHA).		(B)	Quantification	of	
PHA	extracted	from	the	P.	putida	biomass	of	co-cultures	24	h	after	inoculation	into	nitrogen	
replete	(+	N)	or	nitrogen	deplete	(-	N)	medium.	
	has	a	reduced	nitrogen	content	(2mM	NH4Cl)	(Table	3.S1).	Twenty	four	hours	after	cycling	
into	the	nitrogen	deplete	medium,	cell	culture	was	harvested,	dried,	and	processed	for	
quantification	of	methyl	esters	of	alkanoic	acids	by	high-pressure	liquid	chromatography	
(HPLC;	see	Methods).	Processing	of	these	samples	showed	that	both	P.	putida	EM·S	and	P.	
putida	EM·DNT·S	co-cultures	in	nitrogen	deplete	medium	accumulated	4.9	and	5.1	mg	
PHA/L	of	culture,	resulting	in	polymer	contents	of	22	and	23.4	mg	of	PHA/g	cell	dry	weight,	
respectively.		

Discussion	

In	this	work,	we	demonstrate	metabolic	division	of	labor	within	an	artificial	
microbial	co-culture	consisting	of	engineered	strains	of	S.	elongatus	PCC	7942	and	P.	putida	
EM173.	In	this	consortium,	S.	elongatus	CscB	produces	soluble	sugars	using	light	and	CO2	as	
	

117	

	

inputs,	providing	sufficient	organic	compounds	to	promote	the	growth	of	co-cultivated	
strains	of	 P.	putida.	In	turn,	we	demonstrate	that	 P.	putida	simultaneously	degrades	the	
environmental	pollutant	2,4-DNT	while	producing	the	bioplastic	precursor,	PHA.		
	We	have	shown	that	encapsulation	of	 S.	elongatus	CscB	cells	in	alginate	hydrogel	
beads	allowed	the	cyanobacteria	to	persist	in	the	presence	of	2,4-DNT	(Figs.	3.S1-S2	in	the	
Supplemental	Materials)	and	maintain	consistent	production	of	sucrose	after	culture	back	
dilution.	Encapsulated	cyanobacteria	tolerated	concentrations	of	2,4-DNT	5	times	higher	
than	could	be	tolerated	by	planktonic	cyanobacterial	cultures	and	did	so	without	
significantly	changing	chlorophyll	a	content	(Figure	3.2BC).	While	there	is	precedence	for	
the	improved	resilience	of	cells	that	are	encapsulated	within	hydrogels	(Bozeman	et	al.	
1989;	Romo	and	Perez-Martinez	1997;	Weiss	et	al.	2017),	it	is	evident	in	the	literature	that	
even	species	that	share	similar	phototrophic	lifestyles	react	to	hydrogel	encapsulation	very	
differently,	altering	cell	division	rates,	cell	morphology,	and	key	metabolic	pathways	
(Moreno-Garrido	2008).	Increased	resistance	to	environmental	toxins	has	been	attributed	
to	the	physical	properties	of	the	encapsulating	gel	matrices	(e.g.	decreased	diffusion	of	the	
toxin	(Chen	et	al.	1993)),	as	well	as	to	physiological	changes	within	encapsulated	cells	
(Bozeman	et	al.	1989).	The	mechanism	by	which	the	encapsulated	cyanobacteria	resist	the	
effects	of	2,4-DNT	under	these	conditions	has	yet	to	be	elucidated.	Future	work	
investigating	how	this	encapsulation	modulates	cyanobacterium	stress	response	to	
nitroaromatic	compounds	could	yield	genetic	targets	for	modification	to	bolster	the	
resilience	of	the	cyanobacteria	without	the	need	for	mechanical	encapsulation.		
P.	 putida	is	a	gram-negative	soil	bacterium	that	has	recently	gained	significant	
attention	as	a	chassis	for	industrially-relevant	synthetic	biology.	This	is	in	part	thanks	to	the	
	

118	

	

full	sequencing	(Nelson	et	al.	2002)	and	subsequent	generation	of	P.	putida	strains	(e.g.,	P.	
putida	EM173	used	in	this	work)	with	reduced	genomes	that	demonstrate	enhanced	
expression	of	heterologous	proteins	(Martínez-García	et	al.	2015).	 P.	 putida’s	metabolic	
diversity	and	high	tolerance	of	oxidative	stress	make	it	an	ideal	model	organism	for	
studying	toxin	remediation	as	well	as	bioproduction	of	added-value	compounds	(Nikel	et	
al.	2016).	The	introduced	oxidative	2,4-DNT	degradation	pathway	(Figure	3.4A)	is	
chromosomally	integrated	in	this	strain.	While	this	pathway	avoids	generating	highly	
reactive	intermediates	with	hydroxylamino	groups,	the	proteins	in	this	pathway	are	not	yet	
fully	optimized	for	this	new	substrate	(de	las	Heras	et	al.	2011;	Pérez-Pantoja	et	al.	2013),	
leading	to	the	production	of	oxidative	stress	in	the	 Burkholderia	sp.	R34	from	which	the	
pathway	originates	(Pérez-Pantoja	et	al.	2013).	This	oxidative	damage	is	thought	to	
contribute	to	an	increased	rate	of	mutation	and	fosters	a	more	rapid	evolution	of	this	strain	
to	combat	oxidative	stress	(Pérez-Pantoja	et	al.	2013).	While	P.	putida	has	also	been	shown	
to	experience	increased	oxidative	stress	when	actively	utilizing	this	pathway,	this	species	
does	not	exhibit	the	same	rate	of	DNA	damage	and	mutation	(Akkaya	et	al.	2018).	Thus,	the	
physiological	properties	of	P.	putida	allow	for	more	efficient	use	of	an	imperfect	2,4-DNT	
degradation	pathway.	Evolving	or	engineering	this	pathway	toward	increased	specificity	
for	2,4-DNT	could	allow	for	improved	kinetics	and	reduced	ROS	generation.	Conversely,	the	
lower	substrate	specificity	might	allow	this	pathway	to	be	redirected	toward	the	
degradation	of	other	types	of	nitroaromatic	pollutants	from	other	industrial	processes	(Ju	
et	al.	2010).	
2,4-DNT	as	a	compound	represents	a	significant	contaminant	that	poses	a	
bioremediation	challenge.	2,4-DNT	in	its	solid	state	is	very	stable	in	the	environment	as	it	is	
	

119	

	

not	bioavailable,	where	it	often	remains	due	to	its	low	solubility.	Once	solubilized,	2,4-DNT	
can	then	be	reduced	to	highly	reactive	hydroxylamino	intermediates	that	damage	cellular	
machinery	and	DNA	as	chemical	adducts.	Solid	state	2,4-DNT	has	a	high	stability	and	low	
solubility,	allowing	this	compound	to	persist	in	the	environment	and	slowly	disperse	into	
surface	and	ground	water.	Furthermore,	reduced	derivatives	of	this	compound	do	not	
readily	mineralize	in	the	environment	and	persist	for	extended	periods	of	time	in	the	soil	
(Achtnich	et	al.	1999).	The	longevity	of	this	compound	and	its	derivatives	requires	a	long-
term	sustainable	solution	to	remediate	previous	areas	of	contamination,	as	well	as	reduce	
contamination	from	industrial	practices	that	are	actively	producing	2,4-DNT	as	a	by-
product.	Previous	works	that	propose	alternative	biologically	based	methods	of	degrading	
2,4-DNT	are	limited	either	by	the	necessity	to	supply	bioavailable	carbon	(Wang	et	al.	
2011)	or	produce	new	biological	substrate	to	perform	the	degradation	(Oh	et	al.	2016).	
Co-cultures	of	the	2,4-DNT	degrading	 P.	 putida	strain	(P.	 putida	EM·DNT·S)	were	
successfully	grown	while	solely	supported	by	the	fixed	carbon	provided	by	the	
encapsulated	S.	elongatus	CscB.	Furthermore,	these	cultures	were	able	to	successfully	cycle	
over	the	two	weeks	as	a	demonstration	of	this	system’s	stability	and	potential	for	long-
term	sustainable	degradation	of	2,4-DNT.	While	we	demonstrate	intermittent	cycling	of	
media	in	flask-based	cultures	here	(Figure	3.5D),	it	is	possible	to	conceive	of	a	
photobioreactor	that	would	allow	for	continual	introduction	of	2,4-DNT-contaminated	
wastewater.	Here,	the	full	media	exchanges	associated	with	long-term	flask-based	co-
cultures	allowed	us	to	demonstrate	the	ability	of	these	co-cultures	not	only	to	remediate	
this	compound	but	also	produce	the	bioplastic	precursor	polyhydroxyalkanoate	(PHA).	
	

120	

	

While	sustainable	production	of	PHA	has	been	pursued	in	other	contexts,	this	is,	to	
our	knowledge,	the	first	report	in	which	PHA	formation	has	been	concurrent	with	the	
degradation	of	a	xenobiotic	compound.	In	comparison	to	an	independent	report	that	aimed	
to	optimize	PHA	production	from	batch	cultures	of	 S.	 elongatus-P.	 putida	(Löwe	et	al.	
2017a),	we	achieved	a	lower	specific	productivity	in	cultures	simultaneously	degrading	
2,4-DNT.	While	this	current	system	did	not	achieve	the	90%	PHA	of	Cell	Dry	Weight	(CDW)	
value	that	is	commonly	regarded	as	a	benchmark	for	viable	industrial	production	strains,	
production	of	PHA	in	this	system	is	secondary	in	importance	relative	to	the	bioremediation	
of	2,4-DNT.	This	production	of	PHA	served	to	demonstrate	the	potential	diversity	of	
function	inherent	in	this	co-culture	system.	Furthermore,	the	conditions	of	this	system	
could	be	further	optimized	for	PHA	production	by	modifying	total	nitrogen	supplied,	
duration	of	nitrogen	starvation,	or	concentration	of	2,4-DNT.s	Of	note,	we	observed	no	
appreciable	difference	in	P.	putida	growth	rates	between	co-cultures	in	nitrogen	replete	or	
in	low	nitrogen	(2mM	nitrate).	This	raises	the	possibility	that	 P.	 putida	may	be	able	to	
utilize	an	unknown	cyanobacterial	by-product	as	a	nitrogen	source,	and	additional	
optimization	may	be	required	to	fully	activate	PHA	production	pathways.	
One	question	that	arose	as	part	of	this	work	was	whether	the	presence	of	P.	putida	
EM·DNT·S	provides	a	protective	effect	to	S.	elongatus	CscB	in	co-cultures	fed	with	2,4-DNT	
contaminated	medium.	If	this	were	the	case,	it	would	shift	this	relationship	from	a	
commensal	to	a	more	mutualistic	relationship	where	each	species	benefits	from	the	
presence	of	the	other.	Pursuing	longer-term	cultures	with	even	more	rigorous	exposure	to	
2,4-DNT	could	reveal	whether	P.	putida	EM·DNT·S	might	provide	such	a	protective	effect.		
	

121	

	

We	show	that	these	artificial	co-cultures	are	not	only	capable	of	utilizing	media	
contaminated	with	a	toxic	xenobiotic,	but	also	of	producing	the	bioplastic	PHA.	A	scaled	
version	of	this	system	could	hypothetically	take	wastewater	effluent	from	industrial	
sources	contaminated	with	2,4-DNT,	remediate	the	water	allowing	it	to	be	utilized	for	other	
functions,	and	provide	a	mechanism	by	which	PHA	could	be	produced.	The	more	
immediately	relevant	take-away	from	this	work	is	that	photosynthetic	co-cultures	utilizing	
S.	elongatus	CscB	are	flexible	both	in	its	partnerships	with	other	microbes	as	well	as	the	
intended	functionality	of	the	system.	This	creates	opportunities	for	more	advanced	and	
complex	co-cultures	with	new	functions	and	constituents	that	we	hope	will	not	only	
provide	solutions	to	modern	conflicts,	but	also	inform	us	on	how	nature	of	symbiotic	
relationships	develop	in	the	natural	world.	

Methods	

Bacterial	strains	and	culture	conditions	

Planktonic	 S.	 elongatus	strains	were	grown	as	previously	described	(Weiss	et	al.	
2017).	Selection	for	the	genomically-integrated	 cscB	cassette	was	maintained	during	
monoculturing	with	the	addition	of	12.5	ug/mL	chloramphenicol	to	the	medium.	P.	putida	
strains	were	streaked	from	frozen	stocks	onto	selective	LB	plates	and	incubated	at	32°C	
overnight.	The	following	day,	individual	colonies	were	picked	into	LB	media	with	
appropriate	antibiotics.	Plasmid	pSEVA221-cscRABY	in	both	the	 P.	 putida	EM·S	and	 P.	
putida	EM·DNT·S	was	maintained	by	addition	of	50	ug/mL	kanamycin,	while	 P.	 putida	
EM·DNT·S,	carrying	a	genomic	integration	of	the	 dnt	degradation	gene	cluster,	was	
maintained	with	25	μg/mL	gentamicin.	LB	cultures	were	grown	up	overnight	at	32°C	in	a	
	

122	

	

Multitron	(Infors)	incubatory	with	rotary	agitation	at	150	rpm.	The	next	day,	these	were	
utilized	to	inoculate	new	overnight	cultures	in	M3	defined	medium	(see	Table	S1	in	the	
Supplemental	Material)	with	either	20	g/L	or	2	g/L	sucrose,	as	indicated.	These	cultures	
were	then	used	as	inoculum	for	subsequent	experiments.	

 

Encapsulation	of	S.	elongatus	CscB	in	alginate	beads	

Alginate	encapsulation	was	performed	as	previously	described	(Weiss	et	al.	2017)	
with	minor	adjustments.	Briefly,	planktonic	S.	elongatus	CscB	cells	grown	in	BG-11	medium	
+	1g/L	HEPES	(pH	8.3;	Sigma)	were	harvested	at	an	optical	density	measured	at	750	nm	
(OD750)	=	2	via	centrifugation	at	3,500×g	for	30	min	and	concentrated	in	3	mL	of	sulfur-free	
BG-11.	These	cells	were	then	added	to	a	sterile	and	degassed	volume	of	3%	(wt/vol)	
sodium	alginate	and	then	gently	mixed	to	a	final	OD750	=	5.0,	making	a	roughly	2.75%	
(wt/vol)	sodium	alginate-S.	elongatus	CscB	suspension.	In	a	sterile	hood,	this	solution	was	
added	dropwise	to	a	≥20-fold	larger	volume	of	20	mM	BaCl2	using	a	vertically-oriented	
syringe	pump	(KD	Scientific,	Holliston,	MA),	5	mL	syringes	(BD	Biosciences),	and	30	G	
needles.	The	drops	traveled	~35	cm	from	needle	to	the	slowly	stirred	BaCl2	solution	and	
were	allowed	to	cure	in	the	solution	for	at	least	20	min	before	being	rinsed	once	with	BG-
11	medium	and	then	allowed	to	incubate	overnight	in	M3	medium	without	the	additional	
100	mM	NaCl.	This	medium	was	then	exchanged	the	following	day	and	refreshed	again	
with	M3	medium	without	added	NaCl	and	transferred	to	250-mL	baffled	Erlenmeyer	flasks	
and	placed	into	a	Multitron	Pro	(Infors)	Incubator	with	constant	illumination	(15W	Grow-
Lux;	Sylvania;	~70	μmol	m-2	s-1),	2%	CO2	supplementation,	and	shaking	(125	rpm)	at	32°C	
overnight.	The	third	day	post	encapsulation,	the	M3	medium	was	exchanged,	and	the	
	

123	

	

alginate	beads	were	apportioned	in	~10-mL	aliquots	into	125-mL	baffled	Erlenmeyer	
flasks.	The	medium	was	then	replaced	with	M3	+	25	mM	NaCl.	The	concentration	of	salt	
was	then	increased	over	the	next	two	daily	medium	exchanges	to	50	mM	NaCl,	and	then	to	
100	mM	NaCl.	The	medium	was	subsequently	refreshed	daily	for	3-5	days	prior	to	
experiments,	allowing	the	cells	and	alginate	beads	to	fully	stabilize.	In	the	cases	where	the	
beads	would	be	utilized	in	co-culture,	CscB	expression	would	be	induced	the	day	prior	to	
inoculation	of	the	P.	putida	cells,	to	prevent	a	lag	in	protein	expression	from	influencing	
heterotrophic	growth,	and	the	media	exchanged	immediately	prior	to	inoculation.	

Analytical	methods	

Culture	optical	densities	were	measured	with	a	Genesys	20	(Thermo	Fisher	
Scientific,	Waltham,	MA)	spectrophotometer.	Planktonic	S.	elongatus	CscB	was	measured	at	
OD750,	and	both	co-culture	experiments	and	 P.	 putida	monocultures	were	measured	at	
OD600.	This	spectrophotometer	was	also	used	to	measure	chlorophyll	a	concentrations	of	
planktonic	S.	elongatus	cultures	as	in	(Zavřel	et	al.	2015).	For	chlorophyll	a	measurements	
in	alginate-encapsulated	S.	elongatus,	three	technical	replicates	of	four	alginate	beads	were	
placed	in	1.7	mL	eppendorf	tubes.	Then,	1	mL	of	pre-chilled	methanol	was	added	to	the	
beads	which	were	then	gently	vortexed.	These	were	then	incubated	at	4°C	for	30	min	in	the	
dark.	After	briefly	vortexing	the	tubes,	the	solution	containing	the	chlorophyll	 a	was	
removed,	leaving	the	intact	cell	containing	beads	behind.	Absorbance	of	the	extracted	
chlorophyll	 a	was	measured	in	cuvettes	at	720	nm	and	665	nm	to	calculate	the	final	
chlorophyll	a	concentration	as	indicated	elsewhere	(Zavřel	et	al.	2015).	
	

124	

	

Sucrose	concentration	of	beaded	cultures	was	measured	as	indicated	previously	
(Weiss	et	al.	2017)	with	a	slight	modification.	At	select	timepoints,	1	mL	of	culture	
supernatant	was	withdrawn	and	pelleted	at	17,000×g	for	10	min.	The	supernatant	was	
then	transferred	into	a	fresh	tube	for	storage,	3	technical	replicates	of	100	μL	were	
quantified	via	sucrose/D-glucose	assay	kit	(Megazyme,	Bray,	Ireland)	for	each	sample	
(Weiss	et	al.	2017).		
Culture	supernatant	spectra	were	measured	using	a	DU800	Spectrophotometer,	
(Beckman	Coulter,	CA,	USA).	Cell-free	culture	supernatant	was	obtained	by	centrifugation	
for	10	min	at	17,000×g.	The	supernatant	was	then	transferred	to	a	cuvette	for	spectral	
analysis.		GC/MS	and	LC/MS	detection	and	quantitation	were	performed	with	machinery	
housed	in	the	Michigan	State	University	Mass	Spectrometry	and	Metabolomics	Core.	2,4-
DNT	measurements	were	made	using	a	Agilent	5975	GC/single	quadrupole	MS	(Agilent).	
Culture	samples	were	centrifuged	at	17,000×g	for	10	min	and	10	μL	of	cell	free	supernatant	
were	removed	to	a	new	tube.	50	μL	of	ethyl	acetate	was	added	to	the	supernatant	and	
allowed	to	incubate	at	room	temperature	for	30	min.	The	50	μL	of	ethyl	acetate	was	then	
transferred	to	a	GC	vial	and	injected	into	the	GC	machine	(injection	volume	1	μL).	Samples	
were	separated	with	a	5%	phenyl-methyl	capillary	column	(Agilent)	and	measured	by	Mass	
Selective	Detector	(MSD).	A	temperature	of	275°C	was	set	for	the	split/splitless	injector	
(ratio	of	10:1).	Helium	gas	was	used	as	the	carrier	gas	at	a	flow	rate	of	1	mL/min.		
	LC/MS	measurements	were	made	using	a	Waters	Xevo	G2-XS	UPLC/MS/MS	
(Waters).	Culture	samples	were	harvested	and	pelleted	via	centrifugation	(17,000×g	for	10	
min).	A	100-μL	aliquot	of	the	supernatant	was	then	transferred	to	a	clean	tube	and	
	

125	

	

lyophilized.	This	was	then	resuspended	in	1	mL	deionized	water.	Supernatant	samples	
were	harvested	from	culture	and	HPLC	Measurements	of	PHA	accumulation	were	made	
with	a	Waters	2695	HPLC	in	Dr.	Cheryl	Kerfeld’s	lab	at	Michigan	State	University.	Samples	
were	process	and	analyzed	as	previously	described	(Weiss	et	al.	2017).	Briefly,	cells	were	
centrifuged	at	17,000×g	for	10	min,	the	supernatant	decanted,	and	the	pellet	was	
lyophilized.	The	cell	biomass	was	dissolved	in	1	mL	of	concentrated	sulfuric	acid,	heated	to	
90	°C	for	1	h,	cooled	to	room	temperature,	and	then	diluted	100-fold	with	deionized	water.	
A	20-μL	aliquot	was	injected	onto	an	Aminex	300-mm	HPX-87H	(Bio-Rad	Laboratories,	
Hercules,	CA)	column,	and	0.028	N	H2SO4	was	used	as	the	mobile	phase	at	a	flow	of	1	
mL/min.	The	column	temperature	was	maintained	at	60	°C	and	UV-	absorption	was	
monitored	at	210	nm.	Two	standards	of	commercial	PHB	(Sigma-Aldrich)	were	similarly	
treated	and	used	for	quantification	purposes.	

 

	

 

126	

	

	

	

	

	

	

	

APPENDIX	

	

	

127	

	

Appendix:	Chapter	3	Supplemental	Materials	

 

	
Chemical	Name	

Ammonium	iron(III)	citrate	

Calcium	chloride	dihydrate	

Citric	acid	

Cobalt(II)	nitrate	hexahydrate	

Copper(II)	sulfate	pentahydrate	

Dibasic	potassium	phosphate	

Dihydrogen	borate	

Ethylenediaminetetraacetic	acid	(EDTA)	

Magnesium	sulfate	septahydrate	

Manganese	(II)	Chloride	Hydrate	

Sodium	Carbonate	

Sodium	molybdate	dihydrate	

Sodium	nitrate	

Zinc	sulfate	heptahydrate	

	

	

	

	

	

	

	

	

	

	

	

	

	

	

	

	
Chemical	Formula	

C6H8FeNO7	

CaCl2	x	2H2O	

C6H8O7	

Co(NO3)2	x	6H2O	

CuSO4	x	5H2O	

K2HPO4	

H2BO3	

C10H16N2O8	

MgSO4	x	7H2O	

MnCl2	x	4H2O	

Na2CO3	

Na2MoO4	x	2H2O	

NaNO3	

ZnSO4	x	7H2O		

BG-11	

Molarity	

2.29E-05	

2.45E-04	

3.12E-05	

1.70E-07	

3.16E-07	

2.30E-04	

4.70E-05	

3.42E-06	

3.04E-04	

9.15E-06	

1.89E-04	

1.61E-06	

1.76E-02	

7.72E-07	

	

	

	

	

	

	

	

	

	

	

	

	

	

	

	

M3	

Molarity	

2.29E-05	

2.45E-04	

3.12E-05	

1.70E-07	

3.16E-07	

4.73E-03	

4.70E-05	

3.42E-06	

3.04E-04	

9.15E-06	

1.89E-04	

1.61E-06	

1.76E-02	

7.72E-07	

	

	

	

	

	

	

	

	

	

	

	

	

	

	

	

-	

-	

	

	

	

	

	

	

	

	

	

0.1	

NaCl	

NH4Cl	

4.00E-03	

Ammonium	chloride	

Sodium	chloride	
Table	3.S1:	Media	composition	

Chemical	composition	of	BG-11,	M3,	and	M3	-N	media	used	in	this	study.	Between	standard	
BG-11	medium	and	M3	medium	the	only	significant	additions	to	this	media	are:	a	~2	fold	
increase	in	dibasic	potassium	phosphate,	for	buffering	the	pH;	0.1	M	sodium	chloride	to	
stimulate	sucrose	production	in	the	S.	elongatus	CscB;	and	additional	nitrogen	in	the	form	
of	ammonium	chloride,	which	we	have	found	stimulates	the	growth	of	some	heterotrophs	
in	co-culture.	This	additional	nitrogen	is	not	added	in	the	M3	–N	condition	in	which	the	
sodium	nitrate	concentration	is	also	reduced	to	2x10-3	M	or	2	mM	to	trigger	the	
accumulation	of	polyhydroxyalkanoate	(PHA)	in	the	P.	putida	strains.		
	

128	

 

 

	

 
 
 
 
 
 
 
 
 
Figure	3.S1:	Influence	of	2,4-DNT	on	the	growth	of	S.	elongatus	PCC	7942	

(A)	Following	the	growth,	or	lack	thereof,	in	wild	type	cultures	of	S.	elongatus	PCC	7942	
grown	in	BG-11	with	varying	concentrations	of	2,4-DNT	and	inoculated	at	varying	initial	
densities	over	the	course	of	48	hours	(hrs).	The	concentration	of	2,4-DNT	added	to	these	
cultures	corresponds	with	the	estimated	cell	density	as	extrapolated	from	optical	density	
(OD750).	E.g.,	OD750(0.2)	+	62.5uM	2,4-DNT	compared	to	OD750(0.4)	+	120	μM		2,4-DNT.	

 

 

	

129	

	

 

Figure	3.S2:	Encapsulated	S.	elongatus	CscB	in	different	media	with	2,4-DNT	

(A)	Encapsulated	S.	elongatus	CscB	cells	in	BG-11	with	different	added	concentrations	of	
2,4-DNT	[0	μM	-	250	μM]	were	followed	over	the	course	of	6	days.	In	contrast	to	the	
planktonic	cultures	of	S.	elongatus	PCC	7942,	these	cultures	did	not	demonstrate	the	same	
bluing	and	bleaching	transition	associated	with	the	toxic	effects	of	the	2,4-DNT.	(B)	
Encapsulated	 S.	 elongatus	CscB	cells	were	tested	with	this	same	range	of	2,4-DNT	
concentrations	[0	μM	-	250	μM]	in	the	modified	co-culture	media	(M3)	-	100	mM	NaCl,	in	
M3	media	triggering	the	accumulation	of	sucrose	with	the	S.	elongatus	CscB	cells,	and	in	M3	
Media	+	1	mM	IPTG,	the	fully	induced	condition	that	allows	for	the	accumulation	and	
export	of	sucrose	into	the	culture	supernatant.	This	last	condition	is	the	one	utilized	for	all	
co-cultures	unless	otherwise	stated.	Again	no	bleaching	of	the	cells	was	observed.	

	

130	

	

 
Figure	 3.S3:	 Calculated	 chlorophyll	 a	 per	 cell	 comparison	 between	 planktonic	 and	
encapsulated	S.	elongatus	CscB	

(A)	The	calculated	Chl	 a	content	of	planktonic	cells	while	exposed	to	a	range	of	
concentrations	of	2,4-DNT.	The	0	uM	control	cultures	exhibit	a	slight	decrease	of	Chl	a	per	
cell	as	they	are	actively	dividing	over	the	first	24	hours,	diluting	the	existing	Chl	a	between	
daughter	cells.	After	30	hours,	however,	these	cells	begin	to	increase	the	concentration	of	
Chl	 a	as	self-shading	of	the	culture	increases	the	demand	for	light	absorption.	All	of	the	
cultures	that	have	any	concentration	of	2,4-DNT	exhibit	a	steady	decline	in	Chl	a	per	cell	as	
the	cells	react	to	the	2,4-DNT.	(B)	In	order	to	calculate	the	chlorophyll	A	concentration	per	
cell	for	the	alginate	encapsulated	cyanobacteria,	it	was	necessary	to	make	some	
estimations	as	to	the	number	of	beads	in	a	given	flask	as	well	as	the	distribution	of	cells	
within	those	beads.	The	number	of	beads	was	calculated	based	on	the	apparent	diameter	of	
the	alginate	beads	(~2.6	mm)	as	well	as	the	total	volume	of	alginate	beads	in	each	flask	
which	was	established	as	10	mL	of	beads	suspended	in	20	mL	of	medium.	These	numbers	
brought	us	to	the	estimate	of	~1086	beads	per	flask.	In	making	the	alginate	beads,	the	final	
OD750	of	the	cells	resuspended	in	the	alginate	is	~5.0	OD750.	With	an	estimated	3.3x108	S.	
elongatus	PCC	7942	cells	per	OD750	per	1	mL,	we	were	then	able	to	calculate	that	each	bead	
upon	formation	should	have	approximately	1.52x107	cyanobacteria	cells.	As	previous	work	
has	shown	that	almost	no	divisions	occur	within	these	hydrogels	over	an	extended	period	
of	time	(Weiss	et	al.	2017),	we	assumed	this	initial	cell	count	remained	unchanged	for	the	
purposes	of	these	estimates.	From	these	calculations	we	were	then	able	to	give	a	final	
estimate	on	the	Chl	 a	content	per	cell.	Interestingly,	the	per	cell	concentration	of	the	
beaded	cyanobacteria	is	very	much	aligned	with	the	initial	Chl	a	content	of	the	cells	in	the	
planktonic	cultures.	We	believe	this	indicates	these	estimates	are	within	a	reasonable	
range	of	the	actual	per	cell	Chl	a	content.	
	

131	

 

 

	

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Figure	3.S4:	Comparing	P.	putida	supernatant	spectra	over	time	

(Top)	The	P.	putida	EM!DNT!S	spectrum	from	Figure	3	included	for	comparison.	(Bottom)	
The	P.	putida	EM!S	spectrum	from	the	same	experiment	shows	no	significant	accumulation	
of	any	colored	products.	This	further	illustrates	that	the	observed	color	change	is	specific	to	
the	oxidative	pathway	degradation	products.	The	selected	spectra	are	an	average	of	n	=	3.	

 
 
 

	

132	

	

Figure	3.S5:	4M5NC	control	elution	profile	and	m/z	

(Top)	Elution	profile	of	8	uM	4M5NC	standard	with	inset	m/z	chromatograph	of	4M5NC	
peak.	(Middle)	Elution	profile	of	 P.	 putida	EM(cid:0)DNT(cid:0)S	supernatant	elution	profile	with	
inset	m/z	chromatograph	of	4M5NC	peak.	(Bottom)	Solvent	split	for	LC/MS	methods.	
	

133	

 

 

	

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure	 3.S6:	 Loss	 of	 2,4-DNT	 from	 P.	 putida	 monocultures	 and	 reductive	 pathway	
analysis	

(A)	Disappearance	of	2,4-DNT	in	P.	putida	monocultures	grown	in	M3	medium	with	20	g/L	
sucrose	and	250	μM	2,4-DNT	measured	by	GC/MS	(n	=	3,	error	bars	represent	standard	
deviations).	(B)	2,4-DNT	media	controls	were	run	under	both	the	P.	putida	growth		
	

134	

Figure	3.S6	(cont’d)	

conditions	as	well	as	under	the	S.	elongatus/co-culture	conditions.	In	all	cases	there	was	no	
apparent	loss	of	2,4-DNT	by	GCMS	over	the	course	of	48	hrs.	(C)	Reductive	degradation	of	
2,4-DNT	[adapted	from	(Shemer	et	al.	2018)].	Molecules	in	brackets	are	hypothetical.	(D)	
LC/MS	measurements	of	2-amino-4-nitrotoluene/4-amino-2-nitrotoluene	[2A4NT/4A2NT]	
from	the	supernatants	of	monocultures	at	the	4	h	timepoint.	Statistical	analysis	showed	
that	this	decrease	was	statistically	significant	(p-value	=	0.02568)	from	P.	putida	EM!S	to	
the	 P.	 putida	EM!DNT!S	indicating	that	there	is	decreased	accumulation	of	this	first	
measurable	intermediate	of	the	reductive	pathway.	
	

	

	

	

135	

CHAPTER	4:	DIRECTED	EVOLUTION	OF	ESCHERICHIA	COLI	W	CSCR	WITH	ALGINATE	

ENCAPSULATED	SYNECHOCOCCUS	ELONGATUS	CSCB	

The	work	presented	in	this	chapter	is	unpublished.	

	

	
		

136	

	

	

	

Introduction	

How	interspecies	microbial	interactions	arise	and	become	fixed	into	stable	
symbiotic	interactions	in	nature	is	a	fundamental	ecological	question.	Cyanobacteria	are	
particularly	prolific	in	their	establishment	of	mutualistic	interactions	with	a	wide	array	of	
different	organisms,	including	archaea,	other	prokaryotes,	fungi,	plants,	and	animals.	These	
interactions	are	often	shaped	by	millions,	or	even	billions	(e.g.,	chloroplast	endosymbiosis)	
of	years	of	co-evolution	of	the	partner	species.	It	is	challenging	to	deduce	the	origins	of	
symbioses	relying	only	on	observations	of	existing	natural	systems	because	such	
symbioses	are	the	legacies	of	a	history	of	interactions	between	their	ancestral	populations.	
A	pivotal	question	about	the	evolution	of	interacting	microorganisms	is	to	what	extent	
their	metabolic	exchange	networks	emerge	early	as	a	result	of	fortuitous	encounters	
long-term	 stable	
between	 biochemically-compatible	 species,	 versus	 a	 product	 of	
associations	that	increasingly	drive	a	tighter	integration	of	the	communities’	“interactome.”	
In	Chapter	1,	we	discussed	the	natural	interactions	that	have	evolved	between	the	
cyanobacterium	 Prochlorococcus	and	α-proteobacterium	SAR11	as	an	example	of	a	
cyanobacterium/heterotroph	relationship.	Prochlorococcus	has	lost	the	ability	to	mitigate	
ROS	and	due	to	the	high	light	intensity,	excretes	fixed	carbon	as	a	mechanism	of	balancing	
the	cells	redox	state	(Morris	et	al.	2011;	Braakman	et	al.	2017).	SAR11	has	evolved	the	
capacity	to	utilize	that	excreted	carbon	and	improved	its	ROS	mitigation	capacity	
(Giovannoni	2005;	Morris	et	al.	2011;	Zinser	2018).	While	we	can	observe	these	current	
interactions	and	determine	which	factors	are	important	to	maintaining	this	relationship,	it	
is	very	difficult	to	ascertain	the	evolutionary	path	that	eventually	led	to	this	result	due	to	
	

137	

	

the	dynamic	genetic	landscape	in	both	of	these	organisms	and	the	extended	period	of	time	
over	which	they	have	co-evolved.		
One	of	the	primary	advantages	of	utilizing	synthetic	microbial	consortia,	as	
discussed	in	Chapter	1,	is	that	the	interacting	species	in	these	systems	do	not	need	to	have	
a	prior	evolutionary	history	with	one	another,	permitting	an	opportunity	to	study	the	early	
stages	of	evolution	between	symbiotic	partners	from	the	“bottom	up.”	This	is	a	burgeoning	
field	of	research	that	is	still	in	its	infancy,	but	has	already	shown	that	synthetic	consortia	
are	potentially	reflective	of	natural	consortia.	For	example	in	(Hays	et	al.	2017),	 S.	
cerevisiae	W303	and		B.	subtilis	3610	experienced	significant	growth	impairment	at	high	
cyanobacterial	cell	concentrations	while	the	 E.	 coli	W	cscR		thrived	under	the	same	
conditions.	It	was	also	discovered	that	the	 E.	 coli	W	cscR	strain	could	subsist	in	this	
environment	off	of	unknown	component(s)	in	the	cyanobacterial	exudates.	The	differences	
in	performance	under	identical	environmental	conditions	and	with	the	same	phototrophic	
partner	by	these	organisms,	suggests	that	underlying	metabolic	differences	makes	each	
species	more	or	less	compatible	in	a	consortium.	Elucidating	these	interactions	found	
within	these	synthetic	consortia	may	provide	insight	into	important	metabolisms	that	exist	
in	 natural	 consortia.	 Furthermore,	 these	
interactions	 are	 also	 significant	 to	 the	
development	of	productive	synthetic	consortia	and	may	provide	mechanisms	by	which	
bioproduction	from	these	consortia	could	be	enhanced.	However,	disentangling	the	various	
consortium-relevant	metabolic	differences	between	the	three	species	presented	in	this	
example	presents	a	significant	challenge.		
One	method	by	which	we	may	obtain	both	insight	into	metabolic	pathways	relevant	
to	phototrophic	co-culture	as	well	as	potentially	improved	heterotrophic	bioproduciton	
	

138	

	

chassis	is	through	directed	evolution	of	heterotrophic	strains	under	co-culture	conditions.	
Numerous	laboratory	evolution	studies	have	been	conducted	upon	isolated	model	
microbial	species	(Lenski	et	al.	1991;	Boder	et	al.	2000;	Gore	et	al.	2009;	Kawecki	et	al.	
2012;	Ratcliff	et	al.	2012;	Johnson	et	al.	2016;	Pandey	et	al.	2016;	Lenski	2017).	Some	of	
these	experiments	have	been	particularly	impactful	for	theory	development,	as	they	allow	
real-time	observation	of	the	molecular	mechanisms	that	propel	an	organism	across	an	
evolutionary	landscape	towards	higher	fitness	under	a	given	selective	pressure.	One	of	the	
conditions	required	in	directed	experimental	evolution	is	that	the	selective	pressure	on	the	
organisms	remains	consistent.	To	minimize	environmental	variation,	most	instances	of	
experimental	evolution	focus	on	a	single	organism	and	apply	consistent	abiotic	stressors	or	
other	artificial	methods	of	selection	on	that	species	(Lenski	et	al.	1991;	Boder	et	al.	2000;	
Maharjan	et	al.	2006;	Johnson	et	al.	2016;	Tizei	et	al.	2016).		There	are	relatively	few	
examples	of	laboratory	co-evolution	where	selective	pressures	are	applied	to	mixtures	of	
two	or	more	species	and	adaptation	of	both	species	is	looked	at	simultaneously	
(Brockhurst	and	Koskella	2013;	Koskella	and	Brockhurst	2014).	A	common	quandary	of	
analyzing	co-evolved	species	is	that	each	species	has	independent	influence	on	the	
environment,	thereby	shaping	the	selective	pressure.	This	can	give	rise	to	increased	
variability	between	separate	populations	that	are	evolved	under	identical	experimental	
protocols,	and	raises	“chicken	or	egg”	style	questions	about	observed	adaptations.	That	is,	
did	a	mutation	in	species	A	arise	as	result	of	a	change	in	species	B	or	did	it	trigger	the	
change	in	species	B?		Given	that	it	is	difficult	to	verify	causative	adaptive	mutations	even	in	
well-studied	model	organisms	such	as	 E.	 coli	(Lenski	et	al.	1998),	determination	of	the	
selective	forces	shaping	co-evolved	species	can	be	quite	challenging.	
	

139	

	

In	this	chapter,	I	initiated	experiments	designed	to	gain	insight	into	the	dominant	
selective	pressures	facing	the	heterotrophic	species	partnered	with	S.	elongatus	CscB	in	the	
engineered	co-cultures	discussed	previously	(see	Chapter	3).	It	is	preferable	to	avoid	the	
complications	of	co-evolution	of	both	species	mentioned	above	and	it	would	also	be	
possible	that	negative	selection	on	the	sucrose-production	cassette	in	 S.	 elongatus	CscB	
would	decrease	the	stability	of	the	co-culture	in	the	long	term.	The	work	of	Weiss	et	al.	
(2017)	demonstrated	that	 S.	 elongatus	CscB	can	be	encapsulated	in	alginate	hydrogels,	
remaining	viable	and	capable	of	sucrose	export	for	long	time	periods	(>5	months).	Cell	
division	of	the	encapsulated	cyanobacteria	is	dramatically	slowed	(Weiss	et	al.	2017),	
decreasing	the	likelihood	of	genetic	changes	that	would	alter	culture	conditions	
experienced	by	the	partnered	heterotroph.	I	therefore	decided	to	utilize	alginate	
encapsulated	 S.	elongatus	CscB	as	a	platform	with	which	I	could	expose	a	heterotrophic	
strain	to	the	selective	pressures	of	co-cultivation	with	a	naïve	cyanobacterial	strain	for	a	
sufficient	number	of	generations	to	permit	adaptation	of	enhanced	fitness	traits.		

Results	

Selection	and	engineering	of	the	ancestral	strain	

	
It	has	been	previously	shown	that	 E.	coli	W	can	metabolize	sucrose	(Archer	et	al.	
2011;	Sabri	et	al.	2013a)	and	that	it	can	be	stably	co-cultivated	over	long	time	periods	with	
S.	 elongatus	CscB,	subsisting	on	cyanobacterial	exudates	as	the	sole	source	of	organic	
carbon	(Hays	et	al.	2017).		Furthermore,	E.	coli	has	a	well-developed	molecular	toolkit	and	
has	been	routinely	used	for	laboratory	evolution	studies	(Lenski	et	al.	1998;	Yang	et	al.	
2013).	The	 E.	coli	W	cscR	strain	has	had	the	sucrose	catabolism	repressor	gene	deleted	
	

140	

	

from	its	genome,	which	further	enhances	its	utilization	of	sucrose	(Arifin	et	al.	2011).	This	
strain	has	been	previously	co-cultivated	with	S.	elongatus	CscB	for	weeks	in	planktonic	co-
culture	(Hays	et	al.	2017),	therefore	it	was	an	appropriate	background	to	begin	directed	
evolution	studies.	
I	next	incorporated	differential	fluorescent	reporter	tags	into	the	ancestral	strain	
that	could	be	later	used	to	facilitate	the	differentiation	of	the	evolved	and	ancestral	strains	
of	E.	coli	W.	I	believed	that	use	of	fluorescence	would	provide	the	most	utility,	allowing	for	
easy	identification	of	the	strains	by	fluorescent	microscopy	as	well	as	flow	cytometry	for	
quantitative	measurements.	Many	examples	of	this	strategy	can	be	found	in	the	co-culture	
oriented	literature	(Kerner	et	al.	2012;	Tecon	and	Or	2017).	Because	of	the	extra	metabolic	
burden	to	retain	plasmids,	we	opted	to	incorporate	the	fluorescent	protein	expression	
cassettes	into	the	genome	of	our	strain	via	the	method	developed	in	(Sabri	et	al.	2013b).	
This	was	to	increase	genetic	stability	of	expression,	eliminating	the	possibility	of	plasmid	
loss	during	the	experimental	evolution.	We	generated	two	variants	of	the	pKIKOlacZCm	
plasmid,	pKIKOlacZCm-mKOK	and	pKIKOlacZCm-mNG,	encoding	the	orange	mKOK	
(Tsutsui	et	al.	2008)	and	mNeonGreen	(Shaner	et	al.	2013)	proteins,	respectively	(Figure	
4.1A).	Expression	of	these	proteins	is	driven	by	the	constitutive	J23100	promoter,	which	
was	selected	as	the	strongest	driver	of	expression	from	the	Registry	of	Biological	Parts:	
Anderson	Promoter	Collection	(Figure	4.1A).	These	constructs	were	then	transformed	and	
integrated	into	the	 E.	 coli	W	cscR	strains	to	generate	the	base	strains	of	 E.	 coli	W	cscR	
ΔlacZ::mKOK	(E.	coli	W	cscR	mKOK)	and	E.	coli	W	cscR	ΔlacZ::mNeonGreen	(E.	coli	W	cscR	
mNG)	(see	methods).	Fluorescent	microscopy	of	these	strains	shows	uniform	expression	of	
both	proteins	(Figure	4.1B)	though	the	relative	intensity	of	the	mKOK	appears	lower	than	
	

141	

	

that	of	mNeonGreen	which	may	be	due	to	the	difference	in	absorbance	spectra	(515nm	vs	
505nm)	and	quantum	yield	(0.61	vs	0.80)	(Tsutsui	et	al.	2008;	Shaner	et	al.	2013).		Multiple	
clonal	stocks	of	these	strains	were	made	such	that	either	could	be	used	as	the	ancestral	
strain.		

Figure	4.1	E.	coli	W	cscR	lacZ::Fluorophore	lines.	

	
(A)	pKIKO	vector	maps	of	the	plasmids	used	in	this	work	to	genomically	integrate	
expression	cassettes	of	the	fluorophores,	mNeonGreen	(Left)	or	mKOK	(Right).		(B)	
Fluorescent	microscopy	images	of	 E.	coli	W	cscR	lacZ::mNeonGreen	(Left)	and	 E.	coli	W	
cscR	lacZ::mNeonGreen	(Right).	
	

142	

Experimental	selection	conditions	for	photosynthetic	co-culture	

	
Once	the	strains	were	generated,	we	initiated	a	preliminary	trial	of	experimental	
evolution	with	the	 E.	 coli	W	cscR	ΔlacZ::mNeonGreen	strain	to	determine	whether	this	
strain	would	be	able	to	tolerate	co-culture	conditions.	Alginate	beads	containing	 S.	
elongatus	CscB	were	distributed	into	16	flasks	and	inoculated	with	 E.	 coli	W	cscR	
ΔlacZ::mNeonGreen	immediately	after	the	beads	were	fully	cured	(see	Chapter	3	methods).	
These	co-cultures	were	grown	in	phosphate	buffered	BG-11	in	which	we	include	an	
additional	4mM	dibasic	potassium	phosphate	(BG-11P).	These	cultures	were	then	
incubated	in	a	photobioreactor	and	followed	over	the	course	of	the	next	50	days	with	
frequent	serial	dilutions.	Routine	samples	were	collected	to	determine	E.	coli	cell	density	
(via	CFUs)	as	well	as	screen	for	possible	contaminants.	The	culture	medium	was	removed	
and	refreshed	every	3-5	days	and	frozen	stocks	of	the	actively	evolving	cultures	were	made	
at	each	back	dilution	(Figure	4.2A).	This	method	of	experimental	evolution	could	be	
considered	a	hybrid	between	continuous	and	serial	transfer	(Barrick	and	Lenski	2013).	The	
alginate	encapsulated	cyanobacteria	are	constantly	secreting	photosynthetically-derived	
exudates	the	 E.	 coli	can	metabolize	(Hays	et	al.	2017),	similar	to	continuous	culture	or	
bioreactor-based	systems	in	which	fresh	carbon	sources	can	be	continually	added	to	the	
culture.	However	unlike	these	systems,	other	nutrients	(nitrogen,	phosphate,	 etc.,)	can	
become	limited	resources	and	waste	products	accumulate	until	the	medium	is	completely	
refreshed	upon	back	dilution,	analogous	to	conventional	serial	transfer	experiments.	These	
back	dilutions	also	function	as	genetic	bottlenecks	that	randomly	reduce	the	population	
size	and	improve	beneficial	mutation	fixation	rates	(LeClair	and	Wahl	2017).		
	

	

	

143	

	

	

Figure	 4.2:	 E.	 coli	 W	 cscR	 lacZ::mNeonGreen	 experimental	 evolution	 with	 alginate	
encapsulated	S.	elongatus	CscB	

	
(A)	E.	coli	W	cscR	lacZ::mNeonGreen	average	density	in	co-culture	over	the	course	of	the	50	
day	experiment	(n=11	flasks,	5	were	excluded	due	to	contamination).	Culture	density	was	
measured	by	serial	dilution	and	counting	colony	forming	units	(CFUs).	(B)	This	graph	
displays	the	accumulated	total	of	generations	from	the	course	of	this	experiment.	From	
this,	we	are	able	to	calculate	the	estimated	value	of	1.45	generations	per	day,	a	value	in	line	
with	previous	estimates	of	planktonic	co-culture	growth	(Ducat	et	al.	2012).		(C)	Splitting		
	

144	

	

Figure	4.2	(contt’d)	

the	existing	11	flasks	into	group	A	(n=5)	and	group	B	(n	=	4),	we	examined	the	effect	of	
adding	an	additional	4	mM	nitrogen	in	the	form	of	ammonium	chloride	NH4Cl.		
	The	population	of	E.	coli	W	in	the	co-cultures	stably	oscillated	between	an	initial	cell	
concentration	of	N0	=	~107	and	a	final	cell	concentration	NF	=	~109	cells	per	mL	over	the	
course	of	this	experiment.	The	generation	turnover	rate	averaged	~	1.45	generations	per	
day	and	by	the	end	of	this	experiment,	we	had	accumulated	~70	generations	over	the	
course	of	50	days	(Figure	4.2B).		During	the	course	of	our	media	analysis,	we	discovered	
that	E.	coli	W	could	not	efficiently	utilize	the	nitrate	that	is	supplied	as	the	sole	nitrogen	
source	(Archer	et	al.	2011)	in	the	co-culture	buffer.	We	therefore	modified	this	medium	by	
supplementing	4mM	NH4Cl	(M3	medium;	Chapter	3	Methods),	which	increased	the	titer	of	
E.	coli	by	~4	fold	under	co-culture	conditions	(Figure	4.2C).	Higher	total	populations	in	
experimental	evolution	studies	have	been	shown	to	increase	the	overall	pool	of	potential	
mutations,	increasing	rate	beneficial	mutation	fixation	(LeClair	and	Wahl	2017).	
Furthermore,	removing	nitrogen	limitation	as	a	selective	pressure	will	allow	additional	
evolutionary	leeway	to	explore	mutations	related	to	sucrose	utilization	and	ROS-
mitigation.	
Beyond	the	change	in	medium,	there	were	a	few	other	options	to	increase	the	E.	coli	
	
W	growth	rates	in	co-culture.	One	option	was	to	increase	the	dilution	factor	of	co-culture	to	
achieve	higher	generation	turn	over	between	back	dilutions.	The	co-cultures	were	
experiencing	a	roughly	100-fold	reduction	in	population	size	with	each	back	dilution.	
However,	prior	research	that	modeled	the	influence	of	bottlenecks	on	beneficial	mutation	
fixation	rates	indicated	that	the	dilution	factor	plays	a	significant	role	in	random	loss	of	
beneficial	mutations	(Wahl	et	al.	2002).		Wahl	et	al.	calculated	that	a	10-fold	dilution	results	
	

145	

Long-term	co-culture	to	select	for	enhanced	fitness	

in	a	given	beneficial	mutation	having	an	approximate	75%	probability	of	fixation.	
Increasing	this	to	the	commonly	utilized	10,000-fold	dilution	factor	lowered	the	probability	
of	fixation	to	1%	(Wahl	et	al.	2002).	Thus,	attempting	to	achieve	a	higher	generation	count	
between	back	dilutions	would	negatively	impact	the	likelihood	of	beneficial	mutations	
becoming	fixed	in	our	evolving	populations.	We	therefore	instead	opted	to	increase	the	
frequency	of	our	back	dilutions	to	once	daily	to	increase	turnover	in	our	population.		
Making	the	necessary	modifications	to	both	the	media	and	dilution	protocol,	I	
	
initiated	a	refined	evolution	experiment	(Figure	4.3).	14	baffled	culture	flasks	containing	
10	mL	of	alginate	bead	encapsulated	S.	elongatus	CscB	were	inoculated	with	M3	media	with	
~0.5	OD600	E.	coli	W	cscR	lacZ::mNeonGreen.	After	an	initial	5	day	adjustment	period,	these	
co-cultures	were	then	back	diluted	daily	for	the	remainder	of	the	experiment	and	samples	
were	taken	every	7	days	to	make	frozen	stocks.		Looking	at	a	representative	period	of	this	
experiment	between	day	13	and	day	19	of	co-culture,	we	can	see	that	the	E.	coli	cell	density	
is	oscillating	between	N0	=	~107	and	NF	~109	cells/mL	every	24	hours	(Figure	4.3A).	While	
this	is	roughly	the	same	amplitude	observed	in	the	previous	experiment	(Figure	4.2A),	the	
rate	at	which	these	densities	are	achieved	is	significantly	increased,	with	a	turn	over	of	~	
6.7	generations	daily.		This	increased	rate	allowed	for	us	to	more	than	triple	the	number	of	
generations	accumulated	by	the	end	of	this	preliminary	experiment,	~253	generations	
(Figure	4.3B)	as	compared	to	~68	generations	(4.2B).	This	rate	of	growth	translates	to	
roughly	a	3.3	hour	doubling	time,	~	1/3	the	rate	of	 E.	 coli	W	growth	in	rich	medium	
	

	

146	

	

(Shiloach	et	al.	1975)	indicating	that	these	co-cultures	are	performing	well,	but	there	is	still	
room	for	improvement.		

	

Figure	 4.3:	 E.	 coli	 W	 cscR	 lacZ::mNeonGreen	 experimental	 evolution	 with	 daily	
dilutions		

(A)	A	representative	period	of	E.	coli	W	density	from	day	13	to	day	19	of	the	co-cultures.	
CFU’s	were	measured	by	serial	dilution	subsequent	colony	counting.	(B)	Accumulated	
number	of	generations	was	plotted	to	create	this	graph,	from	which	a	linear	trend	line	
could	be	used	to	project	future	generation	numbers	as	the	experiment	progressed.	The	
average	number	of	generations	per	day	increased	to	6.7	and	the	average	doubling	time	was	
3.3	hours.		
	

Preliminary	verification	of	enhanced	fitness	in	evolved	lines	

One	of	our	hypotheses	was	that	consistent	but	diffuse	levels	of	sucrose	produced	by	
the	 S.	elongatus	CscB	would	place	a	significant	selective	pressure	on	the	 E.	coli,	favoring	
mutations	involved	in	improved	sucrose	uptake	and/or	catabolism.	I	therefore	compared	
the	capacity	of	ancestral	and	evolved	strains	to	grow	in	defined	media	at	decreasing	
sucrose	content.	To	establishing	a	baseline	for	sucrose	utilization,	we	first	performed	side-
by-side	
inoculations	 of	 the	 ancestral	 and	 evolved	 strains	 of	 E.	 coli	 W	 cscR	
lacZ::mNeonGreen	at	~0.4	OD600	into	15	mL	culture	tubes	filled	with	M3	medium	+	20	g/L	
sucrose	(Figure	4.4A).		As	noted	earlier,	 E.	 coli	W	cscR	is	naturally	able	to	catabolize	
sucrose	from	the	environment.	Thus,	in	medium	with	20	g/L	of	sucrose	shows	the	ancestral	
	

147	

	

strain	and	evolved	strains’	optical	densities	increasing	at	similar	rates.	However,	when	the	
sucrose	concentration	was	decreased	to	1.25	g/L	in	a	secondary	experiment	in	which	the	

Figure	 4.4:	 Preliminary	 evidence	 for	 enhanced	 fitness	 of	 evolved	 strains	 in	 low	
sucrose	medium	

(A)	Baseline	growth	of	both	the	ancestral	and	evolved	strains	demonstrated	the	ability	to	
utilize	sucrose	as	a	carbon	source.	This	experiment	was	performed	in	M3	medium	+	20	g/L	
sucrose.	Strains	are	listed	along	the	right	side	of	the	graph	in	order	of	highest	OD600	at	the	
last	measured	time	point.	(B)	Lowering	the	concentration	of	sucrose	to	1.25	g/L	
demonstrated	that	a	significant	number	of	the	evolved	strains	had	improved	fitness	
relative	to	the	ancestral	strain.	(A	&	B)	These	experiments	were	performed	in	15	mL	
culture	tubes	with	n	=	3.		
	

	

148	

	

strains	were	inoculated	at	0.1	OD600,	the	growth	rates	of	the	evolved	strains	and	the	
ancestral	strain	begin	to	diverge	(Figure	4.4B).	These	results	suggest	that	over	the	course	
of	250	generations	the	majority	of	the	experimentally	co-evolved	strains	may	have	
acquired	mutations	that	enable	them	to	better	utilize	low	concentrations	of	sucrose.	

Discussion	

In	this	chapter	I	present	my	preliminary	data	on	the	long-term	(>250	generations)	
directed	evolution	of	 E.	 coli	W	cscR	in	co-culture	with	alginate	bead	encapsulated	 S.	
elongatus	CscB.	Utilizing	a	fluorescently	labeled	E.	coli	W	cscR	strain,	I	performed	proof-of-
concept	experiments	to	evolve	heterotrophic	strains	better	adapted	to	growing	in	the	low	
concentrations	of	sucrose	produced	in	phototrophic	co-culture	with	 S.	 elongatus	CscB.	
While	this	work	describes	very	preliminary	testing	of	the	evolved	E.	coli	W	strains	the	fact	
that	we	have	observed	significantly	enhanced	growth	at	low	sucrose	concentrations	is	
promising	data	to	support	this	system	as	a	viable	approach	to	experimental	evolution.	
Encapsulating	the	S.	elongatus	CscB	in	alginate	severely	limits	cell	division	(Weiss	et	
al.	2017),	significantly	reducing	the	likelihood	that	 S.	 elongatus	would	spontaneously	
mutate	in	a	way	that	eliminates	sucrose	export	(e.g.,	a	cscB	null	mutation).	Furthermore,	
this	method	of	isolation	allows	the	cells	to	interact	with	the	supernatant	as	well	as	allows	
for	facile	removal	of	the	planktonic	heterotrophs	without	disrupting	the	cyanobacterial	
population.	This	facilitates	the	maintenance	of	consistent	cyanobacterial	based	inputs	into	
the	co-cultures	(e.g.,	sucrose	and	other	exudates),	therefore	stabilizing	associated	selective	
pressures	for	the	heterotroph	over	time.		
	

149	

	

	
We	hypothesized	that	two	driving	forces	in	this	experimental	evolution	would	be	
the	hyperoxia	and	low	abundance	of	carbon.	This	approach	may	generate	E.	coli	mutants	
with	enhanced	fitness	while	growing	in	co-culture	with	cyanobacteria	through	multiple	
mechanisms.	For	example,	we	would	expect	increased	fitness	in	strains	with	mutations	that	
enhance	utilization	of	sucrose	or	other	secreted	products	from	S.	elongatus.	Alternatively,	
E.	 coli	mutants	with	increased	resistance	to	hyperoxia	may	also	be	recovered.	At	high	
cyanobacterial	densities	in	co-culture,	they	can	inhibit	the	growth	and	viability	of	the	
partner	heterotroph	species.	This	effect	is	density	dependent,	is	only	observed	in	the	light,	
and	can	result	in	instability	or	extinction	of	the	heterotrophic	microbe	at	high	
cyanobacterial	densities	(Hays	et	al.	2017).	This	effect	may	be	multifaceted,	but	we	strongly	
suspect	hyperoxygenation	and/or	ROS	production	caused	by	oxygenic	photosynthesis	
contributes	to	the	heterotrophic	viability	loss.	Additionally,	other	limited	micronutrients	
that	are	only	refreshed	upon	back	dilution	of	the	medium	(phosphate,	iron,	etc.,)	may	also	
be	a	source	of	selective	pressure.	Nitrogen	limitation	may	have	been	a	significant	selective	
pressure	during	the	initial	round	of	experimental	evolution	as	there	was	a	drastic	
improvement	in	 E.	 coli	growth	after	transitioning	to	the	ammonium	supplemented	M3	
medium.	Interestingly,	this	also	revealed	that	E.	coli	W	was	able	to	catabolize	an	as	of	yet	
unidentified	cyanobacterial	exudate	as	a	source	of	nitrogen	which	may	be	a	subject	of	
future	investigation.	
	
Moving	forward	with	this	work,	genome	sequencing	of	the	evolved	strains	can	be	
used	to	identify	mutations	conferring	fitness	advantages.	It	is	commonly	accepted	that	the	
mutation	rate	of	a	given	microbe	is	strongly	correlated	with	the	overall	size	of	its	genome	
with	a	calculated	genomic	mutation	rate	of	~0.0033	per	DNA	replication	(Drake	1991).		
	

150	

	

This	implies	that	per	every	1000	replications	we	would	expect	to	see	~3	mutants	arise.	
However,	our	co-culture	environment	may	be	providing	additional	mutagenic	stress	in	the	
form	of	ROS	that	has	the	potential	to	increase	the	rate	of	mutation	in	the	E.	coli	strains.	In	
conditions	in	which	the	mutation	rate	is	relatively	low,	a	beneficial	mutation	that	arises	
often	does	so	in	isolation	and	undergoes	a	“selection	sweep”	in	which	this	lineage	quickly	
increases	in	frequency	to	dominate	the	culture	(Desai	and	Fisher	2007).		This	property	
would	facilitate	identification	of	beneficial	mutations	within	the	evolved	cultures	as	we	
would	expect	these	mutations	to	rise	to	fixation	quickly.	It	should	also	be	noted	that	other	
examples	of	experimental	evolution	with	 E.	 coli	witnessed	the	emergence	of	beneficial	
mutations	in	fewer	than	50	generations	(Marchal	et	al.	2017).	
Future	analyses	of	these	evolved	lines	will	include	the	quantification	of	relative	
fitness	advantage	over	the	ancestral	strain	in	direct	flask-based	competitions,	via	flow	
cyotometry	analysis,	mediated	by	the	genomically	encoded	mNeonGreen	and	mKOK	
fluorophores.	Subsequent	full-genome	sequencing	of	these	generated	strains	will	allow	for	
the	identification	of	causative	beneficial	mutations.	Areas	of	particular	interest	will	be	
those	in	which	a	number	of	the	independently	generated	lines	converged	on	beneficial	
mutations	in	specific	genes	or	pathways.	These	mutations	can	then	be	recapitulated	in	
unevolved	laboratory	strains	to	verify	whether	these	mutations	do	have	a	positive	fitness	
effect	under	co-culture	conditions.	Additionally,	by	making	regular	frozen	stocks	of	these	
evolving	lines,	we	can	trace	the	temporal	origination	of	these	mutations	through	the	
strains.	Ultimately,	the	lines	generated	by	this	approach	may	be	used	to	gain	“bottom	up”	
insight	into	early	adaptations	that	occur	between	microbes	and	cyanobacteria	that	increase	
	

151	

	

compatibility.	Because	of	the	modularity	of	this	platform,	there	is	the	potential	to	examine	
multiple	heterotrophic	species	in	this	manner	and	determine	common	pathways	that	
mutate	in	ways	important	for	adapting	to	a	commensal	lifestyle	that	is	dependent	upon	
cyanobacterial	primary	productivity.	

Materials	and	Methods	

Bacterial	strains,	media	and	growth	conditions	

S.	elongatus	CscB	as	cultured	and	alginate	encapsulated	as	described	in	the	Chapter	
3	Methods.	E.	coli	W	cscR	strains	were	regularly	cultured	in	LB	medium	and	were	struck	on	
rich	medium	LB	plates	from	frozen	stocks	and	grown	at	32°C	in	an	Multitron	Pro	(Infors).	
E.	coli	W	cscR	fluorescent	lines	were	generated	as	in	(Sabri	et	al.	2013b).	

Co-culture	conditions	

Co-culture	evolution	of	the	E.	coli	strains	were	performed	in	125-mL	baffled	flasks	
containing	10	mL	of	alginate	encapsulated	S.	elongatus	CscB	and	20	mL	of	M3	medium	with	
1	mM	IPTG	(Chapter	3).	These	cultures	were	grown	in	a	Multitron	Pro	(Infors)	incubator	
with	constant	illumination	by	fluorescent	bulbs	(15W	Gro-Lux;	Sylvania;	~70	uml	m-2	S-1),	
at	32°C,	with	supplemented	2%	CO,	and	rotary	shaking	at	125	rpm.	E.	coli	cell	density	was	
regularly	checked	via	spot	platting	serial	dilutions	(dilution	range	10-2-10-7)	of	the	culture	
onto	LB	medium	agarose	plates,	which	were	allowed	to	grow	in	an	incubator	overnight	at	
37°C.	To	generate	frozen	stocks	of	evolving	cultures,	2	mL	samples	of	the	culture	was	taken	
and	the	cells	were	pelleted	via	centrifugation	at	17,000×g	for	10	mins.	These	cells	were	
then	resuspended	in	a	mixture	of	LB	and	30%	glycerol	and	frozen	at	-80°C.	Upon	back	
	

152	

	

dilution,	the	entirety	of	the	20	mL	of	culture	was	removed	from	the	baffled	culture	flasks,	
leaving	the	alginate	beads	in	place.	The	flasks	and	beads	were	then	rinsed	with	20	mL	of	
fresh	M3	medium.	After	briefly	shaking	the	flask,	this	medium	was	also	removed	and	a	
fresh	M3	medium	was	added	back	into	the	culture.	1	mM	IPTG	was	then	added	to	maintain	
the	CscB	expression	of	the	alginate	encapsulated	S.	elongatus	CscB.	

Imaging	and	spectrophotometery	

These	technologies	and	techniques	were	performed	as	previously	described	in	

Chapter	2	and	Chapter	3.	
	

	

	

153	

	

CHAPTER	5:	CONCLUSIONS	AND	FUTURE	PROSPECTIVES	

	

154	

	

	

	

Overview	

This	dissertation	works	on	developing	tools	and	functions	to	enable	artificial	microbial	
consortia	that	include	the	photosynthetic	cyanobacterium	 S.	 elongatus	PCC	7942.	In	
Chapter	2.1,	we	created	a	functional	surface	display	system	utilizing	a	recombinant	
internally	tagged	version	of	the	endogenous	outer	membrane	protein,	SomA.	We	
demonstrate	that	this	surface	display	system	allows	these	cyanobacteria	to	bind	to	
functionalized	abiotic	beads	and	engineered	S.	cerevisiae	(Fedeson	and	Ducat	2016).	I	then	
show	preliminary	data	in	Chapter	2.2	where	I	diversify	the	peptides	displayed	by	the	
cyanobacteria	and	engineer	both	E.	coli	W	cscR	and	S.	cerevisiae	strains	to	express	surface	
bound	protein	domains	to	facilitate	interspecies	adhesion.	Chapter	3	details	our	work	in	
creating	a	bioremediating	co-culture	of	the	sucrose	exporting	strain	of	 S.	elongatus	CscB	
and	the	2,4-DNT	degrading	 P.	putida	Suc+DNT+	strain.	Encapsulation	of	the	 S.	elongatus	
CscB	cells	within	alginate	hydrodel	beads	allows	the	cyanobacteria	to	endure	the	2,4-DNT	
and	maintain	excretion	of	sucrose	to	support	the	P.	putida.	Engineered	to	express	a	suite	of	
enzymes	that	specifically	degrade	2,4-DNT	via	an	oxidative	pathway,	measurement	of	
colored	intermediates	indicated	that	the	engineered	strain	was	indeed	oxidizing	the	2,4-
DNT	and	lowering	the	flux	through	the	more	damaging	reductive	pathway.	Furthermore,	
the	P.	putida	also	accumulated	the	bioplastic	precursor	PHA	when	cultures	were	cycled	into	
nitrogen	deplete	media.	Finally,	in	Chapter	4,	we	perform	a	directed	evolution	experiment	
with	co-cultured	 E.	 coli	W	cscR	and	alginate	encapsulated	 S.	 elongatus	CscB.	Once	
optimized,	this	experiment	achieved	over	250	generations	of	 E.	 coli	W	cscR	in	14	
independently	evolved	flasks,	with	preliminary	competition	results	demonstrating	that	the	
evolved	strains	were	outcompeting	the	ancestral	strain	in	a	sucrose	only	minimal	medium.		
	

155	

	

Here,	I	discuss	some	of	the	broader	questions	and	implications	related	to	my	work	and	the	
potential	of	synthetic	microbial	consortia.	

Diffuse	resources	and	spatial	structure	

An	important	component	of	cooperative	microbial	interactions	that	rely	on	diffusive	
public	goods	relates	to	how	these	microbes	maintain	partner	specificity	(Allen	et	al.	2013).	
These	mechanisms	include	physical	adhesion	or	local	association	(Momeni	et	al.	2013;	
Tecon	and	Or	2017)	within	a	community,	the	exclusion	or	inhibition	of	competing	species,	
and	interspecies	communication	and	translocation	(Liaimer	et	al.	2015).	In	this	
dissertation,	I	have	experimented	with	two	approaches	applicable	to	artificial	consortia.	
The	first	involved	creating	a	biological	mechanism	by	which	cyanobacteria	could	be	
induced	to	establish	direct	physical	association	with	appropriately	functionalized	surfaces	
(Chapter	2).	The	second	approach	utilized	alginate	encapsulation	to	maintain	a	constant	
population	of	cooperating	cyanobacteria	to	provide	fixed	carbon	for	heterotrophic	species	
(Chapter	3	and	Chapter	4).	Both	systems	highlight	the	importance	of	spatial	organization;	
one	focused	on	bringing	the	species	together,	whereas	the	other	emphasized	the	benefits	of	
keeping	the	interacting	species	separate.	
Spatial	association	is	often	used	to	overcome	challenges	associated	with	diffusible	
metabolite	exchange.	For	example,	experimental	evolution	of	 S.	cerevisiae	W303	to	grow	
solely	on	low	concentrations	of	diffuse	sucrose	resulted	in	the	evolution	of	division	
mutants	that	allowed	daughter	cells	to	remain	adhered	to	the	mother	cell,	increasing	the	
localized	uptake	of	the	resource	by	related	cells	(Koschwanez	et	al.	2013).	Similarly,	
another	study	utilized	auxotrophic	strains	of	E.	coli	that	engaged	in	passive	cross-feeding,	
	

156	

	

allowing	the	two	strains	to	grow	in	culture	without	supplementation.	Over	the	course	of	16	
serial	transfers	of	these	co-cultures,	mutations	arose	that	resulted	in	aggregation	of	the	
cross-feeding	strains	in	co-culture	(Marchal	et	al.	2017).	Convergence	on	adhesion	in	these	
experimental	evolution	studies	supports	the	idea	that	spatial	structuring	of	cooperative	
communities	may	be	an	important	consideration	in	the	development	of	synthetic	consortia.	
This	position	is	further	strengthened	by	the	observations	that	spatial	organization	of	
consortia	favors	the	cooperating	species,	which	provide	positive	feedback	to	culture	
growth	(Chuang	et	al.	2009;	Rossetti	et	al.	2010;	Momeni	et	al.	2013).	The	work	of	Rossettie	
et	al.	modeled	the	evolution	of	cheater	cells	arising	in	unicellular	and	multicellular	
cyanobacteria,	 and	
in	 the	 aggregate/multicellular	 populations,	 these	
differentiated	cells	were	unable	to	outcompete	the	cooperating	vegetative	cells	without	
risking	extinction.	Thus,	subdividing	the	population	into	structured	aggregates,	limits	the	
evolution	or	invasion	of	cheating	cell	types.		

found	 that	

Natural	examples	of	structured	consortia	

These	experiments	echo	the	many	existing	associations	that	can	be	found	in	
interactions	 between	
naturally	 occurring	 symbioses	 and	 consortia.	 Even	
Prochlorococcus	and	SAR11	(discussed	in	Chapter	1),	which	are	often	separated	by	100	or	
more	cell	lengths	(Biller	et	al.	2015)	in	the	highly	turbulent	surface	layer	of	the	open	ocean,	
appear	to	have	some	level	of	spatial	organization,	with	Prochlorococcus	ecotypes	stratifying	
in	a	niche-dependent	manner	(Johnson	et	al.	2017).	Future	research	into	this	stratification	
may	reveal	additional	features	of	community	structure,	potentially	reflected	in	the	specific	
“helper”	microbes	(e.g.,	SAR11)	that	dominate	a	given	stratum.	If	the	association	between	
	

the	

157	

	

Prochlorococcus	and	SAR11	ecotypes	were	to	represent	a	relatively	unstructured	and	
distant	spatial	organization,	the	other	end	of	the	spectrum	of	microbial	interactions	with	
high	levels	of	physical	and	structural	association	could	be	represented	by	the	holobiont,	
Chlorochromatium	 aggregatum	(Liu	et	al.	2013).	This	consortium	consists	of	the	motile	
heterotrophic	bacterium	 Candidatus	 Symbiobacter	 mobilis,	with	~15	cells	of	adhered	
photolithoautotrophic	 bacterium	 Chlorobium	 chlorochromatii.	 The	 tight	 association	
between	these	two	species	has	allowed	for	massive	reductions	in	the	Ca	S.	mobilis	genome,	
rendering	it	entirely	dependent	on	its	epibionts	and	no	longer	capable	of	independent	
growth	(Liu	et	al.	2013).	This	kind	of	genome	reduction	is	reminiscent	of	that	observed	in	
plant	and	algal	plastids	(Howe	et	al.	2008).	This	lends	credence	to	the	hypothesis	that	the	
ancestral	eukaryote	and	cyanobacteria	may	have	also	had	a	similarly	strong	physical	
association	prior	to	the	primary	endosymbiotic	event.		Current	evidence	suggests	that	
another	important	evolutionary	transition,	the	colonization	of	land	by	plant	life,	was	likely	
aided	by	interactions	with	symbiotic	fungi	(Delaux	et	al.	2015).	Thus,	the	preadaptation	of	
the	ancestral	eukaryotic	and	cyanobacterial	species	is	an	area	of	research	that	warrants	
further	investigation	to	generate	a	more	complete	understanding	of	one	of	the	most	
important	evolutionary	events	of	all	time.	While	we	are	able	to	make	some	inferences	
through	phylogenetic	analysis	of	modern	photosynthetic	microorganisms,	such	as	that	of	
the	primary	endosymbiont	likely	belonging	to	the	Gloeomargaritales	(Ponce-Toledo	et	al.	
2017),	it	is	nearly	impossible	to	accurately	predict	the	nature	of	the	environment	and	
interactions	that	would	have	allowed	for	the	primary	endosymbiosis	to	occur	solely	from	
phylogenetic	data.	Additionally,	although	we	are	able	to	look	at	contemporary	examples	of	
natural	eukaryotic	and	cyanobacterial	symbioses	for	clues	regarding	what	traits	might	
	

158	

	

facilitate	endosymbiotic	events,	as	discussed	in	Chapter	1,	it	is	often	times	difficult	to	
disentangle	highly	integrated	relationships.	This	and	other	similar	avenues	of	research	may	
benefit	from	the	utilization	of	tailored	synthetic	consortia.	

Artificial	consortia	and	modularity	

The	usage	of	naïve	and	genetically	tractable	organisms	to	explore	fundamental	
questions	such	as,	“How	do	bacteria	adapt	to	having	photosynthetic	neighbors?	(Chapter	
4)”	allow	for	a	significant	level	of	control	while	sacrificing	a	variable	amount	of	realism	
depending	on	the	questions	being	pursued	(Figure	1.1).	However,	direct	co-cultures	are	not	
always	suitable	for	every	area	of	research.	In	Chapter	3,	we	found	it	necessary	to	
encapsulate	our	sucrose	exporting	cyanobacteria	in	alginate	hydrogels	not	only	to	maintain	
consistent	sucrose	production,	but	also	to	allow	the	cyanobacteria	to	withstand	the	toxic	
effects	of	2,4-DNT	in	the	co-culture	medium.	Furthermore,	other	examples	of	preliminary	
consortia	have	shown	that	some	species	are	more	or	less	compatible	than	others	(Hays	et	
al.	2017).	One	pertinent	example	is	that	of	a	tripartite	culture	of	 Azotobacter	vinelandii,	
Bacillus	licheniformis,	and	Paenibacillus	curdlanolyticus	that	failed	to	thrive	in	isolation	or	in	
direct	association	were	able	to	survive	when	grown	in	distantly	connected	microfluidics	
wells	survived	(Kim	et	al.	2008).	Because	this	area	of	research	is	still	developing,	there	is	a	
lack	of	guiding	principles	to	direct	the	construction	of	these	synthetic	consortia.	Each	new	
combination	of	species	assembled	into	a	consortium	risks	unanticipated	metabolic	
interactions	stymieing	the	consortium’s	productivity	or	even	triggering	its	collapse.	Thus,	
to	enable	the	eventual	expansion	of	potential	species	combinations,	work	is	being	done	to	
	

159	

	

characterize	or	create	strains	to	act	as	modular	components	that	can	be	interchanged	
between	consortia	(Figure	1.2).		
The	pursuit	of	microbial	strains	with	the	potential	for	“plug-and-play”	like	
functionality	in	artificial	consortia	is	based	on	the	same	view	that	led	to	the	creation	of	
BioBricks™,	the	belief	that	by	generating	standardized	modules,	future	researchers	will	
have	the	diversity	of	components	with	which	to	design	and	explore	consortia	with	new	and	
unique	functions.	While	my	own	dissertation	work	with	consortia	has	been	limited	to	
utilizing	strains	of	E.	coli,	P.	putida,	and	S.	elongatus,	others	have	begun	to	explore	a	number	
of	other	species	like	Azotobacter	vinelandii	(Ortiz-Marquez	et	al.	2014;	Smith	and	Francis	
2016,	2017)	with	potential	to	become	modular	components.		

Alginate	encapsulation	as	a	technique	for	artificial	consortia	

In	both	Chapter	3	and	Chapter	4,	the	co-culture	approach	utilized	entailed	the	
encapsulation	of	the	S.	elongatus	CscB	strain	within	alginate	hydrogels.	Chapter	3	served	as	
a	direct	demonstration	of	how	applied	synthetic	ecology	and	rationally	designed	co-
cultures	could	be	utilized	to	address	a	pressing	concern	in	our	society.	Chapter	4	focused	
on	the	experimental	evolution	of	 E.	 coli	in	a	co-culture	environment	in	which	the	only	
source	of	carbon	was	the	cyanobacterial	photosynthates	from	the	alginate	encapsulated	S.	
elongatus	CscB	strain.	The	reasoning	behind	using	this	system	in	each	of	these	projects	is	
addressed	in	their	respective	chapters.	However,	there	are	additional	features	of	this	
system	relevant	to	the	development	of	synthetic	consortia	that	warrant	discussion.	One	of	
the	primary	features	of	this	system	is	that	it	physically	separates	the	planktonic	microbial	
species	from	the	encapsulated	species	(in	these	cases,	the	cyanobacterium).	While	similar	
	

160	

	

in	some	respects	to	culture	techniques	that	harness	two	chambers	separated	by	
membranes	filter	(Paul	et	al.	2013;	Moutinho	et	al.	2017)	or	cultures	separated	by	dialysis	
tubes/bags	(Paul	et	al.	2009),	these	techniques	do	not	maintain	a	fixed	population	as	has	
been	shown	in	the	alginate	hydrogel	beads	(Weiss	et	al.	2017).	Cell	density	and	growth	
phase	can	drastically	influence	how	cells	interact	with	their	surroundings.	This	especially	
true	in	photosynthetic	organisms	like	cyanobacteria	that	often	regulate	gene	expression	
based	on	circadian	rhythms	(Markson	et	al.	2013;	Piechura	et	al.	2017)	or	in	microbes	that	
utilize	density	dependent	signaling	mechanisms	like	quorum	sensing	(Waters	and	Bassler	
2005).	The	use	of	alginate	encapsulated	cells	thus	acts	as	another	level	of	control,	reducing	
some	of	the	variables	related	to	the	encapsulated	species	growth	at	the	cost	of	exploring	
how	changing	the	ratio	between	species	influences	consortia	stability.	Approaching	this	co-
culture	method	with	modularity	in	mind,	one	of	the	potential	benefits	is	that	it	allows	for	
the	rapid	one-sided	testing	of	the	encapsulated	species	on	the	planktonically	grown	
species.	Additionally,	this	influence	is	likely	to	be	mediated	by	diffusible	exudates	which	
narrows	the	number	of	factors	that	need	to	be	accounted	for	when	trying	to	identify	the	
causative	metabolite(s).	

The	intersection	of	synthetic	ecology	and	experimental	evolution	

My	final	discussion	point	focuses	on	how	the	intersection	of	rationally	designed	
consortia	and	experimental	evolution	may	present	researchers	with	a	unique	tool	to	
approach	difficult	questions.	As	alluded	to	earlier	in	this	chapter,	I	believe	that	questions	
surrounding	species	adaptation	to	one	another,	especially	in	the	context	of	symbiotic	
relationships,	could	be	approached	with	a	combination	of	these	techniques.	When	we	
	

161	

	

identify	a	system	or	metabolic	pathway	that	we	believe	is	important	for	a	symbiotic	
interaction,	often	times	one	of	the	first	ways	in	which	we	attempt	to	understand	it	is	by	
removing	or	altering	its	components.	By	methodically	perturbing	the	system,	we	can	
establish	what	pathways	are	critical,	redundant,	or	unnecessary.	Furthermore,	if	that	
system	shares	homology	with	mechanisms	in	other	related	species	we	can	also	begin	to	
determine	how	it	has	evolved	over	time.	However,	it	is	difficult	to	put	these	pieces	back	
into	their	evolutionary	context	to	explore	how	the	interaction	between	these	symbiotic	
organisms	shaped	their	co-evolution.	With	experimental	evolution	of	synthetic	consortia,	
as	opposed	to	removing	functions	from	an	existing	symbiosis,	you	could	instead	work	from	
the	“bottom-up”	by	adding	rationally	defined	functions	(like	cell	adhesion)	to	a	species	in	a	
co-culture	and	then	observe	how	this	change	influences	the	development	of	these	strains	as	
they	co-evolve.		
The	work	presented	here	has	significant	potential	to	further	increase	awareness	of	
and	interest	in	cyanobacteria	as	well	as	synthetic	microbial	ecology.	At	the	time	of	this	
writing,	a	second	researcher	group	unaffiliated	with	my	work	has	published	a	manuscript	
generating	a	surface	display	system	similar	to	the	one	I	describe	in	Chapter	2,	for	the	
cyanobacterium	Synechocystis	sp.	6803	(Cengic	et	al.	2018).	It	is	my	hope	that	others	will	
continue	to	explore	the	potential	of	both	cyanobacteria	and	artificial	consortia.			
	

	

	

162	

	

	

	

	

	

	

	

	

REFERENCES	

163	

	

	

	

	
	

REFERENCES	

Abramson	BW,	Kachel	B,	Kramer	DM,	Ducat	DC.	Increased	photochemical	efficiency	in	
cyanobacteria	via	an	engineered	sucrose	sink.	Plant	Cell	Physiol.	2016;57(12):2451–
60.		
	Achtnich	C,	Sieglen	U,	Knackmuss	HJ,	Lenke	H.	Irreversible	binding	of	biologically	reduced	
2,4,6-trinitrotoluene	to	soil.	Environ	Toxicol	Chem.	1999;18(11):2416–23.		
	Adams.	Symbiotic	Interactions.	Springer;	2000.		
	Adams	DG,	Bergman	B,	Nierzwicki-Bauer	SA,	Duggan	PS,	Rai	AN,	Schüßler	A.	
Cyanobacterial-plant	symbioses.	Prokaryotes	Prokaryotic	Biol	Symbiotic	Assoc.	2013.	
p.	359–400.		
	Aikens	J,	Turner	RJ.	Transgenic	photosynthetic	microorganisms	and	photobioreactor.	2013.	
p.	231pp.		
	Akkaya	Ö,	Pérez-pantoja	DR,	Calles	B,	Nikel	PI,	Lorenzo	V	de.	The	metabolic	redox	regime	of	
Pseudomonas	putida	tunes	its	evolvability	towards	novel	xenobiotic	substrates.	MBio.	
2018;9:e01512-18.		
	Allen	B,	Gore	J,	Nowak	MA.	Spatial	dilemmas	of	diffusible	public	goods.	Elife.	2013;2013(2).		
	Altschul	SF,	Gish	W,	Miller	W,	Myers	EW,	Lipman	DJ.	Basic	Local	Alignment	Search	Tool.	J	
Mol	Biol.	1990;215:403–10.		
	Ambrosio	R,	Ortiz-Marquez	JCF,	Curatti	L.	Metabolic	engineering	of	a	diazotrophic	
bacterium	improves	ammonium	release	and	biofertilization	of	plants	and	microalgae.	
Metab	Eng.	2017;40:59–68.		
	Amin	SA,	Green	DH,	Hart	MC,	Kupper	FC,	Sunda	WG,	Carrano	CJ.	Photolysis	of	iron-
siderophore	chelates	promotes	bacterial-algal	mutualism.	Proc	Natl	Acad	Sci.	
2009;106(40):17071–6.		
	Andreu	C,	Del	Olmo	M.	Yeast	arming	by	the	Aga2p	system:	effect	of	growth	conditions	in	
galactose	on	the	efficiency	of	the	display	and	influence	of	expressing	leucine-
containing	peptides.	Appl	Microbiol	Biotechnol.	2013;97(20):9055–69.		
	Andrianantoandro	E,	Basu	S,	Karig	DK,	Weiss	R.	Synthetic	biology:	new	engineering	rules	
for	an	emerging	discipline.	Mol	Syst	Biol.	2006;2:1–14.		
	Archer	CT,	Kim	JF,	Jeong	H,	Park	JH,	Vickers	CE,	Lee	SY,	et	al.	The	genome	sequence	of	E.	coli	
W	(ATCC	9637):	comparative	genome	analysis	and	an	improved	genome-scale	
reconstruction	of	E.	coli.	BMC	Genomics.	2011;12:9.		
	

164	

	

	Arifin	Y,	Sabri	S,	Sugiarto	H,	Krömer	JO,	Vickers	CE,	Nielsen	LK.	Deletion	of	cscR	in	
Escherichia	coli	W	improves	growth	and	poly-3-hydroxybutyrate	(	PHB	)	production	
from	sucrose	in	fed	batch	culture.	J	Biotechnol.	2011;156(4):275–8.		
	Azam	F,	Malfatti	F.	Microbial	structuring	of	marine	ecosystems.	Nat.	Rev.	Microbiol.	2007.	p.	
782–91.		
	Bailliez	C,	Largeau	C,	Casadevall	E,	Chimie	L	De,	Physique	O,	Nationale	E.	Applied	o	.	d	
Microbiology	Biotechnology	Growth	and	hydrocarbon	production	of	Botryococcus	
braunii	immobilized	in	calcium	alginate	gel.	1985;99–105.		
	Bais	HP,	Weir	TL,	Perry	LG,	Gilroy	S,	Vivanco	JM.	The	role	of	root	exudates	in	rhizosphere	
interations	with	plants	and	other	organisms.	Annu	Rev	Plant	Biol.	2006;57:233–66.		
	Bar-On	YM,	Phillips	R,	Milo	R.	The	biomass	distribution	on	Earth.	Proc	Natl	Acad	Sci.	
2018;201711842.		
	Barrick	JE,	Lenski	RE.	Genome	dynamics	during	experimental	evolution.	Nat	Rev	Genet.	
2013;14:827–39.		
	Barry	A,	Wolfe	A,	English	C,	Ruddick	C,	Lambert	D.	2016	National	Algal	Biofuels	Technology	
Review.	Bioenergy	Technol.	Off.	2016.		
	Basu	S,	Gerchman	Y,	Collins	CH,	Arnold	FH,	Weiss	R.	A	synthetic	multicellular	system	for	
programmed	pattern	formation.	Nature.	2005;434(7037):1130–4.		
	Baumdicker	F,	Hess	WR,	Pfaffelhuber	P.	The	infinitely	many	genes	model	for	the	
distributed	genome	of	bacteria.	Genome	Biol	Evol.	2012;	4:(4).443–56.		
	Bergman	B,	Ran	L,	Adams	DG.	Cyanobacterial-plant	Symbioses:	Signaling	and	Development.	
Cyanobacteria	Mol	Biol	Genomics,	Evol.	2008.	447–73.		
	Bertilsson	S,	Berglund	O,	Pullin	MJ,	Chisholm	SW.	Release	of	dissolved	organic	matter	by	
Prochlorococcus.	Vie	Milieu.	2005;55(3–4):225–31.		
	Besingi	RN,	Clark	PL.	Extracellular	protease	digestion	to	evaluate	membrane	protein	cell	
surface	localization.	Nat	Protoc.	2015;10(12):2074–80.		
	Biller	SJ,	Berube	PM,	Lindell	D,	Chisholm	SW.	Prochlorococcus:	The	structure	and	function	
of	collective	diversity.	Nat	Rev	Microbiol.	2015;13(1):13–27.		
	Biller	SJ,	Coe	A,	Chisholm	SW.	Torn	apart	and	reunited:	Impact	of	a	heterotroph	on	the	
transcriptome	of	Prochlorococcus.	ISME	J.	2016;10(12):2831–43.		
	Biller	SJ,	Coe	A,	Roggensack	SE,	Chisholm	SW.	Heterotroph	Interactions	Alter	
	

165	

	

Prochlorococcus	Transcriptome	Dynamics	during	Extended	Periods	of	Darkness.	
mSystems.	2018;3(3):e00040-18.		
	Blasche	S,	Kim	Y,	Oliveira	AP,	Patil	KR.	Model	Microbial	Communities	for	Ecosystems	
Biology.	Curr	Opin	Syst	Biol.	2017;(September):1–7.		
	van	Bloois	E,	Winter	RT,	Kolmar	H,	Fraaije	MW.	Decorating	microbes:	Surface	display	of	
proteins	on	Escherichia	coli.	Trends	Biotechnol.	2011;29(2):79–86.		
	Boder	ET,	Midelfort	KS,	Wittrup	KD.	Directed	evolution	of	antibody	fragments	with	
monovalent	femtomolar	antigen-binding	affinity.	Proc	Natl	Acad	Sci	USA.	
2000;97(20):10701–5.		
	Boder	ET,	Wittrup	KD.	Yeast	surface	display	for	screening	combinatorial	polypeptide	
libraries.	Nat	Biotechnol.	1997;15(6):553–7.		
	Boder	ET,	Wittrup	KD.	Yeast	surface	display	for	directed	evolution	of	protein	expression,	
affinity,	and	stability.	Methods	Enzymol.	2000	Jan;328:430–44.		
	Bohannan	BJM,	Lenski	RE.	The	relative	importance	of	competition	and	predation	varies	
with	productivety	in	a	model	community.	Am	Nat.	2000;156(4):329–40.		
	Bork	P,	Bowler	C,	De	Vargas	C,	Gorsky	G,	Karsenti	E,	Wincker	P.	Tara	Oceans	studies	
plankton	at	Planetary	scale.	Science.	2015.	p.	873.		
	Bozeman	J,	Koopman	B,	Bitton	G.	Toxicity	testing	using	immobilized	algae.	Aquat	Toxicol.	
1989;14:345–52.		
	Braakman	R,	Follows	MJ,	Chisholm	SW.	Metabolic	evolution	and	the	self-organization	of	
ecosystems.	Proc	Natl	Acad	Sci.	2017;114(15):E3091–100.		
	Brenner	K,	You	L,	Arnold	FH.	Engineering	microbial	consortia :	a	new	frontier	in	synthetic	
biology.	Trends	Biotechnol.	2008;26(9):483–9.		
	Brockhurst	MA,	Koskella	B.	Experimental	coevolution	of	species	interactions.	Trends	Ecol	
Evol	[Internet].	Elsevier	Ltd;	2013;28(6):367–75.		
	Brussaard	CPD,	Bidle	KD,	Pedrós-Alió	C,	Legrand	C.	The	interactive	microbial	ocean.	Nat.	
Microbiol.	2016;	2		
	Bryant	C,	McElroy	WD.	Nitroreductases.	In:	Muller	F,	editor.	Chem	Biochem	Flavoenzymes,	
Vol	II.	1st	ed.	Boca	Raton,	Florida,	USA:	CRC	Press,	Inc.;	1991.	p.	291–9.		
	Cameron	DE,	Bashor	CJ,	Collins	JJ.	A	brief	history	of	synthetic	biology.	Nat	Rev	Microbiol.	
2014;12(5):381–90.		
	
	

166	

	

Carini	P,	Steindler	L,	Beszteri	S,	Giovannoni	SJ.	Nutrient	requirements	for	growth	of	the	
extreme	oligotroph	“Candidatus	Pelagibacter	ubique”	HTCC1062	on	a	defined	medium.	
ISME	J.	2013;7(3):592–602.		
	Carney	LT,	Lane	TW.	Parasites	in	algae	mass	culture.	Front.	Microbiol.	2014.		
	Carrieri	D,	Paddock	T,	Maness	P-C,	Seibert	M,	Yu	J.	Photo-catalytic	conversion	of	carbon	
dioxide	to	organic	acids	by	a	recombinant	cyanobacterium	incapable	of	glycogen	
storage.	Energy	Environ	Sci.	2012;5(11):9457.		
	Castenholz	RW.	Culturing	Methods	for	Cyanobacteria.	Methods	Enzymol.	1988;167:68–93.		
	Cengic	I,	Uhlén	M,	Hudson	EP.	Surface	display	of	small	affinity	proteins	on	Synechocystis	sp.	
strain	PCC	6803	mediated	by	fusion	to	the	major	type	IV	pilin	PilA1.	J	Bacteriol.	
2018;200(16):1–19.		
	Chaijarasphong	T,	Savage	DF.	Sequestered:	Design	and	Construction	of	Synthetic	
Organelles.	Synth	Biol	Parts,	Devices	Appl.	John	Wiley	&	Sons;	2018;8.		
	Chaparro	JM,	Sheflin	AM,	Manter	DK,	Vivanco	JM.	Manipulating	the	soil	microbiome	to	
increase	soil	health	and	plant	fertility.	Biol.	Fertil.	Soils.	2012.	489–99.		
	Chen	D,	Lewandowski	Z,	Roe	F,	Surapaneni	P.	Diffusivity	of	Cu2+	in	Calcium	Alginate	Gel	
Beads.	Biotechnol	Bioeng.	1993;41:755–60.		
	Chisti	Y.	Biodiesel	from	microalgae.	Biotechnol	Adv.	2007;25(3):294–306.		
	Chisti	Y.	Constraints	to	commercialization	of	algal	fuels.	J	Biotechnol.	Elsevier	B.V.;	
2013;167(3):201–14.		
	Cho	DH,	Ramanan	R,	Heo	J,	Lee	J,	Kim	BH,	Oh	HM,	et	al.	Enhancing	microalgal	biomass	
productivity	by	engineering	a	microalgal-bacterial	community.	Bioresour	Technol.	
2015;175:578–85.		
	Chodkowski	JL,	Shade	A.	A	new	synthetic	community	system	for	probing	microbial	
interactions	driven	by	exometabolites.	mSystems.	2017;2(6):4–7.		
	Chuang	JS,	Rivoire	O,	Leibler	S.	Simpson’s	paradox	in	a	synthetic	microbial	system.	Science.	
2009;323(5911):272–5.		
	Chungjatupornchai	W,	Fa-aroonsawat	S.	Translocation	of	green	fluorescent	protein	to	
cyanobacterial	periplasm	using	ice	nucleation	protein.	J	Microbiol.	2009;47(2):187–92.		
	Chungjatupornchai	W,	Fa-Aroonsawat	S.	Biodegradation	of	organophosphate	pesticide	
using	recombinant	cyanobacteria	with	surface-	and	intracellular-expressed	
organophosphorus	hydrolase.	J	Microbiol.	2008;18(5):946–51.		
	

167	

	

	Chungjatupornchai	W,	Kamlangdee	A,	Fa-Aroonsawat	S.	Display	of	organophosphorus	
hydrolase	on	the	cyanobacterial	cell	surface	using	synechococcus	outer	membrane	
protein	a	as	an	anchoring	motif.	Appl	Biochem	Biotechnol.	2011;164(7):1048–57.		
	Connon	SA,	Giovannoni	SJ.	High-throughput	methods	for	culturing	microorganisms	in	very-
low-nutrient	media	yield	diverse	new	marine	isolates.	Appl	Environ	Microbiol.	
2002;68(8):3878–85.		
	Cooper	MB,	Smith	AG.	Exploring	mutualistic	interactions	between	microalgae	and	bacteria	
in	the	omics	age.	Curr.	Opin.	Plant	Biol.	2015.	147–53.		
	Costello	EK,	Stagaman	K,	Dethlefsen	L,	Bohannan	BJM,	Relman	DA.	The	application	of	
ecological	theory	toward	an	understanding	of	the	human	microbiome.	Science.	
2012;336(6086):1255–62.		
	Coyte	KZ,	Schluter	J,	Foster	KR.	The	ecology	of	the	microbiome:	Networks,	competition,	and	
stability.	Science.	2015;350(6261):663–6.		
	Delaux	P-M,	Radhakrishnan	G	V.,	Jayaraman	D,	Cheema	J,	Malbreil	M,	Volkening	JD,	et	al.	
Algal	ancestor	of	land	plants	was	preadapted	for	symbiosis.	Proc	Natl	Acad	Sci.	
2015;112(43):201515426.		
	Demmig-Adams	B,	Stewart	JJ,	Burch	TA,	Adams	WW.	Insights	from	placing	photosynthetic	
light	harvesting	into	context.	J	Phys	Chem	Lett.	2014;5(16):2880–9.		
	Desai	MM,	Fisher	DS.	Beneficial	mutation-selection	balance	and	the	effect	of	linkage	on	
positive	selection.	Genetics.	2007;176(3):1759–98.		
	Ding	M-Z,	Song	H,	Wang	E-X,	Liu	Y,	Yuan	Y-J.	Design	and	construction	of	synthetic	microbial	
consortia	in	China.	Synth	Syst	Biotechnol.	2016;6–11.		
	Dismukes	GC,	Carrieri	D,	Bennette	N,	Ananyev	GM,	Posewitz	MC.	Aquatic	phototrophs:	
efficient	alternatives	to	land-based	crops	for	biofuels.	Curr	Opin	Biotechnol.	
2008;19(3):235–40.		
	Dodds	WK,	Gudder	DA,	Mollenhauer	D.	THE	ECOLOGY	OF	NOSTOC.	J	Phycol.	1995;31(1):2–
18.		
	Doebeli	M,	Hauert	C.	Models	of	cooperation	based	on	the	Prisoner’s	Dilemma	and	the	
Snowdrift	game.	Ecol	Lett.	2005;8(7):748–66.		
	Dolinšek	J,	Goldschmidt	F,	Johnson	DR.	Synthetic	microbial	ecology	and	the	dynamic	
interplay	between	microbial	genotypes.	FEMS	Microbiol	Rev.	2016;(January):1–19.	
	Drake	JW.	A	constant	rate	of	spontaneous	mutation	in	DNA-based	microbes.	Proc	Natl	Acad	
	
168	

	

Sci.	1991;88(16):7160–4.	
	Du	W,	Liang	F,	Duan	Y,	Tan	X,	Lu	X.	Exploring	the	photosynthetic	production	capacity	of	
sucrose	by	cyanobacteria.	Metab	Eng	.	2013;19:17–25.	
	Duan	Y,	Luo	Q,	Liang	F,	Lu	X.	Sucrose	secreted	by	the	engineered	cyanobacterium	and	its	
fermentability.	J	Ocean	Univ	China.	2016;15(5):890–6.	
	Ducat	DC.	Synthetic	biology:	molecular	tools	for	engineering	organisms.	eLS.	2017		
	Ducat	DC,	Avelar-Rivas	JA,	Way	JC,	Silver	PA.	Rerouting	carbon	flux	to	enhance	
photosynthetic	productivity.	Appl	Environ	Microbiol.	2012;78(8):2660–8.	
	Ducat	DC,	Way	JC,	Silver	PA.	Engineering	cyanobacteria	to	generate	high-value	products.	
Trends	Biotechnol	.	2011;29(2):95–103.	
	Dufresne	A,	Ostrowski	M,	Scanlan	DJ,	Garczarek	L,	Mazard	S,	Palenik	BP,	et	al.	Unraveling	
the	genomic	mosaic	of	a	ubiquitous	genus	of	marine	cyanobacteria.	Genome	Biol.	
2008;9(5).		
	Eiteman	MA,	Lee	SA,	Altman	E.	A	co-fermentation	strategy	to	consume	sugar	mixtures	
effectively.	J	Biol	Eng.	2008;2.		
	Fedeson	DT,	Ducat	DC.	Cyanobacterial	surface	display	system	mediates	engineered	
interspecies	and	abiotic	binding.	ACS	Synth	Biol.	2016;6(2):367–74.		
	Ferri	S,	Nakamura	M,	Ito	A,	Nakajima	M,	Abe	K,	Kojima	K,	et	al.	Efficient	surface-display	of	
autotransporter	proteins	in	cyanobacteria.	Algal	Res.	2015;12:337–40.	
	Franzosa	EA,	Hsu	T,	Sirota-Madi	A,	Shafquat	A,	Abu-Ali	G,	Morgan	XC,	et	al.	Sequencing	and	
beyond:	Integrating	molecular	“omics”	for	microbial	community	profiling.	Nat.	Rev.	
Microbiol.	2015.	p.	360–72.		
	French	CE,	Rosser	SJ,	Bruce	NC.	Biotransformations	of	Explosives.	Biotechnol	Genet	Eng	
Rev.	2001;18(1):171–217.	
	Fuhrman	JA,	Cram	JA,	Needham	DM.	Marine	microbial	community	dynamics	and	their	
ecological	interpretation.	Nat.	Rev.	Microbiol.	2015.	p.	133–46.		
	Gestel	J	VaN,	Vlamakis	H,	Kolter	R.	Division	of	Labor	in	Biofilms :	the	Ecology	of	Cell	
Differentiation.	Microbiol	Spectr	Am	Soc	Microbiol.	2015;(May):1–24.		
	Ghirardi	ML.	Implementation	of	photobiological	H2	production:	The	O2	sensitivity	of	
hydrogenases.	Photosynth.	Res.	2015.	p.	383–93.		
	Gibson	DG,	Young	L,	Chuang	R-Y,	Venter	JC,	Hutchison	C	a,	Smith	HO.	Enzymatic	assembly	
	
169	

	

of	DNA	molecules	up	to	several	hundred	kilobases.	Nat	Methods.	2009	May;6(5):343–
5.	
	Gillet	F,	Roisin	C,	Fliniaux	MA,	Dubreuil	AJ,	Barbotin	JN.	Immobilization	of	Nicotiana	
tabacum	plant	cell	suspensions	within	calcium	alginate	gel	beads	for	the	production	of	
enhanced	amounts	of	scopolin.	2000;26:229–34.		
	Giovannoni	SJ.	Genome	Streamlining	in	a	Cosmopolitan	Oceanic	Bacterium.	Science	(80-	).	
2005;309(5738):1242–5.		
	Giovannoni	SJ.	SAR11	Bacteria:	The	Most	Abundant	Plankton	in	the	Oceans.	Ann	Rev	Mar	
Sci.	2017;9(1):231–55.		
	Glass	DS,	Riedel-Kruse	IH.	A	Synthetic	Bacterial	Cell-Cell	Adhesion	Toolbox	for	
Programming	Multicellular	Morphologies	and	Patterns.	Cell.	2018;174(3):649–
658.e16.		
	Goers	L,	Freemont	P,	Polizzi	KM.	Co-culture	systems	and	technologies:	taking	synthetic	
biology	to	the	next	level.	J	R	Soc	Interface.	2014;11(96):20140065–20140065.		
	Golden	SS,	Brusslan	J,	Haselkorn	R.	[12]	Genetic	engineering	of	the	cyanobacterial	
chromosome.	Methods	Enzymol.	1987;153(C):215–31.		
	Goldford	JE,	Lu	N,	Bajic	D,	Estrela	S,	Tikhonov	M,	Sanchez-Gorostiaga	A,	et	al.	Emergent	
Simplicity	in	Microbial	Community	Assembly.	bioRxiv.	2017;474(August):205831.		
	Gopal	M,	Gupta	A,	Thomas	G	V.	Bespoke	microbiome	therapy	to	manage	plant	diseases.	
Front	Microbiol.	2013;4(DEC).		
	Gore	J,	Youk	H,	van	Oudenaarden	A.	Snowdrift	game	dynamics	and	facultative	cheating	in	
yeast.	Nature.	2009	May	14;459(7244):253–6.		
	Greenfield	LK,	Whitfield	C.	Synthesis	of	lipopolysaccharide	O-antigens	by	ABC	transporter-
dependent	pathways.	Carbohydr	Res.	2012	Jul	15;356:12–24.		
	Griest	WH,	Stewart	AJ,	Vass	AA,	Ho	CH.	Chemical	and	Toxicological	Characterization	of	
Slurry	Reactor	Biotreatment	of	Explosives-Contaminated	Soils.pdf.	Oak	Ridge,	
Tennessee,	USA;	1998.		
	Griest	WH,	Tyndall	RL,	Stewart	AJ,	Caton	JE,	Vass	AA,	Ho	C	-H,	et	al.	Chemical	
characterization	and	toxicological	testing	of	windrow	composts	from	explosives-
contaminated	sediments.	Environ	Toxicol	Chem.	1995;14(1):51–9.		
	Grima	EM,	Belarbi	E,	Fernández	FGA,	Medina	AR,	Chisti	Y.	Recovery	of	microalgal	biomass	
and	metabolites:	process	options	and	economics.	Biotechnol	Adv.	2003;20(7):491–
515.	
	

170	

	

	Gründel	M,	Scheunemann	R,	Lockau	W,	Zilliges	Y.	Impaired	glycogen	synthesis	causes	
metabolic	overflow	reactions	and	affects	stress	responses	in	the	cyanobacterium	
Synechocystis	sp.	PCC	6803.	Microbiol	(United	Kingdom).	2012;158(12):3032–43.		
	Gupta	PL,	Lee	SM,	Choi	HJ.	A	mini	review:	photobioreactors	for	large	scale	algal	cultivation.	
World	J	Microbiol	Biotechnol.	2015;31(9):1409–17.		
	Hagan	CL,	Silhavy	TJ,	Kahne	D.	β-Barrel	membrane	protein	assembly	by	the	Bam	complex.	
Annu	Rev	Biochem.	2011	Jan;80:189–210.		
	Haigler	BE,	Nishino	SF,	Spain	JC.	Biodegradation	of	4-Methyl-5-Nitrocatechol	by	
Pseudomonas	sp.	Strain	DNT.	J	Bacteriol.	1994;176(11):3433–7.		
	Hansel	A,	Pattus	F,	Ju	UJ,	Tadros	MH.	Cloning	and	characterization	of	the	genes	coding	for	
two	porins	in	the	unicellular	cyanobacterium	Synechococcus	PCC	6301.	
1998;1399:31–9.		
	Hansel	A,	Tadros	MH.	Characterization	of	two	pore-forming	proteins	isolated	from	the	
outer	membrane	of	Synechococcus	PCC	6301.	Curr	Microbiol.	1998;36(6):321–6.		
	Hanski	I,	Hanski	I.	Metapopulation	dynamics.	Nature.	1998;396(6706):41–9.		
	Hartmann	M,	Gomez-Pereira	P,	Grob	C,	Ostrowski	M,	Scanlan	DJ,	Zubkov	M	V.	Efficient	CO2	
fixation	by	surface	Prochlorococcus	in	the	Atlantic	Ocean.	ISME	J.	2014;8(11):2280–9.		
	Hays	SG,	Ducat	DC.	Engineering	cyanobacteria	as	photosynthetic	feedstock	factories.	
Photosynth	Res.	2015;123(3):285–95.	
	Hays	SG,	Patrick	WG,	Ziesack	M,	Oxman	N,	Silver	PA.	Better	together:	Engineering	and	
application	of	microbial	symbioses.	Curr	Opin	Biotechnol.	2015;36:40–9.	
	Hays	SG,	Yan	LLW,	Silver	PA,	Ducat	DC.	Synthetic	Photosynthetic	Consortia	Define	
Interactions	Leading	to	Robustness	and	Photoproduction.	J	Biol	Eng.	2017;11(4):1–36.	
	Hickman	JW,	Kotovic	KM,	Miller	C,	Warrener	P,	Kaiser	B,	Jurista	T,	et	al.	Glycogen	synthesis	
is	a	required	component	of	the	nitrogen	stress	response	in	Synechococcus	elongatus	
PCC	7942.	Algal	Res.	2013;2(2):98–106.		
	Holyoak	M,	Lawler	SP.	Persistence	of	an	extinction-prone	predator-prey	interaction	
through	metapopulation	dynamics.	Ecology.	1996;77(6):1867–79.		
	Hom	EFY,	Murray	AW.	Plant-fungal	ecology.	Niche	engineering	demonstrates	a	latent	
capacity	for	fungal-algal	mutualism.	Science.	2014	Jul	4;345(6192):94–8.	
	Hooker	BS,	Skeen	RS.	Transgenic	phytoremediation	blasts	onto	the	scene.	Nat	Biotechnol.	
	

171	

	

1999;17(May):1999.		
	Howe	CJ,	Barbrook		a	C,	Nisbet	RER,	Lockhart	PJ,	Larkum		a	WD.	The	origin	of	plastids.	
Philos	Trans	R	Soc	Lond	B	Biol	Sci.	2008	Aug	27;363(1504):2675–85.	
	Inoue	H,	Nojima	H,	Okayama	H.	High	efficiency	transformation	of	Escherichia	coli	with	
plasmids.	Gene.	1990;96(1):23–8.	
	Ispolatov	I,	Ackermann	M,	Doebeli	M.	Division	of	labour	and	the	evolution	of	
multicellularity.	Proc	R	Soc	B	Biol	Sci.	2012;279(1734):1768–76.		
	Ives	AR,	Carpenter	SR.	Stability	and	diversity	of	ecosystems.	Science.	2007.	p.	58–62.		
	Jarmander	J,	Gustavsson	M,	Do	TH,	Samuelson	P,	Larsson	G.	A	dual	tag	system	for	facilitated	
detection	of	surface	expressed	proteins	in	Escherichia	coli.	Microbial	Cell	Factories;	
2012;11(1):1.	
	Jessup	CM,	Forde	SE,	Bohannan	BJM.	Microbial	Experimental	Systems	in	Ecology.	Adv	Ecol	
Res.	2005;37(04):273–307.		
	Jessup	CM,	Kassen	R,	Forde	SE,	Kerr	B,	Buckling	A,	Rainey	PB,	et	al.	Big	questions,	small	
worlds:	Microbial	model	systems	in	ecology.	Trends	Ecol.	Evol.	2004.	p.	189–97.		
	Johnson	TJ,	Halfmann	C,	Zahler	JD,	Zhou	R,	Gibbons	WR.	Increasing	the	tolerance	of	
filamentous	cyanobacteria	to	next-generation	biofuels	via	directed	evolution.	Algal	
Res.	2016;18:250–6.	
	Johnson	ZI,	Zinser	ER,	Coe	A,	McNulty	NP,	Malcolm	E,	Woodward	S,	et	al.	Niche	Partitioning	
among	Prochlorococcus	Ecotypes	along	Ocean-Scale	Environmental	Gradients.	
Science.	2017;311(5768):1737–40.		
	Jose	J,	Bernhardt	R,	Hannemann	F.	Functional	display	of	active	bovine	adrenodoxin	on	the	
surface	of	E.	coli	by	chemical	incorporation	of	the	[2Fe-2S]	cluster.	ChemBioChem.	
2001;2(9):695–701.		
	Jose	J,	Bernhardt	R,	Hannemann	F.	Cellular	surface	display	of	dimeric	Adx	and	whole	cell	
P450-mediated	steroid	synthesis	on	E.	coli.	J	Biotechnol.	2002;95(3):257–68.		
	Jose	J,	Zangen	D.	Autodisplay	of	the	protease	inhibitor	aprotinin	in	Escherichia	coli.	
Biochem	Biophys	Res	Commun.	2005;333(4):1218–26.		
	Ju	K-S,	Parales	R,	E.	Nitroaromatic	Compounds,	from	Synthesis	to	Biodegradation.	Microbiol	
Mol	Biol	Rev.	2010;74(2):250–72.	
	Kang	HJ,	Baker	EN.	Intramolecular	isopeptide	bonds:	Protein	crosslinks	built	for	stress?	
Trends	Biochem	Sci.	2011;36(4):229–37.	
	
172	

	

	Kashtan	N,	Roggensack	SE,	Rodrigue	S,	Thompson	JW,	Biller	SJ,	Coe	A,	et	al.	Single-Cell	
Genomics	Reveals	Hundreds	of	Coexisting	Subpopulations	in	Wild	Procholococcus.	
Science.	2014;344(April):416–20.	
	Kato	S,	Haruta	S,	Cui	ZJ,	Ishii	M.	Stable	Coexistence	of	Five	Bacterial	Strains	as	a	Cellulose-
Degrading	Community.	Appl	Environ	Microbiol.	2005;71(11):7099–106.		
	Kawecki	TJ,	Lenski	RE,	Ebert	D,	Hollis	B,	Olivieri	I,	Whitlock	MC.	Experimental	evolution.	
Trends	Ecol	Evol.	2012;27(10):547–60.	
	Kazamia	E,	Aldridge	DC,	Smith	AG.	Synthetic	ecology	-	A	way	forward	for	sustainable	algal	
biofuel	production?	J	Biotechnol.	2012a;162(1):163–9.		
	Kazamia	E,	Czesnick	H,	Nguyen	TT	Van,	Croft	MT,	Sherwood	E,	Sasso	S,	et	al.	Mutualistic	
interactions	between	vitamin	B12-dependent	algae	and	heterotrophic	bacteria	exhibit	
regulation.	Environ	Microbiol.	2012b;14(6):1466–76.		
	Kazamia	E,	Helliwell	KE,	Purton	S,	Smith	AG.	How	mutualisms	arise	in	phytoplankton	
communities:	building	eco-evolutionary	principles	for	aquatic	microbes.	Ecol.	Lett.	
2016.	p.	810–22.		
	Kelsic	ED,	Zhao	J,	Vetsigian	K,	Kishony	R.	Counteraction	of	antibiotic	production	and	
degradation	stabilizes	microbial	communities.	Nature.	2015;521(7553):516–9.		
	Kerner	A,	Park	J,	Williams	A,	Lin	XN.	A	programmable	scherichia	coli	consortium	via	
tunable	symbiosis.	PLoS	One.	2012;7(3):1–10.		
	Kerr	B,	Riley	MA,	Feldman	MW,	Bohannan	BJM.	Local	dispersal	promotes	biodiversity	in	a	
real-life	game	of	rock-paper-scissors.	Nature.	2002;418(6894):171–4.		
	Kim	HJ,	Boedicker	JQ,	Choi	JW,	Ismagilov	RF.	Defined	spatial	structure	stabilizes	a	synthetic	
multispecies	bacterial	community.	Proc	Natl	Acad	Sci	U	S	A.	2008;105(47):18188–93.		
	Kinoshita	S.	Amino-acid	Production	by	the	Fermentation	Process.	Nature.	1972.	p.	211.		
	Kiran	B,	Kumar	R,	Deshmukh	D.	Perspectives	of	microalgal	biofuels	as	a	renewable	source	
of	energy.	Energy	Convers	Manag.	2014;88:1228–44.	
	Kondo		a,	Ueda	M.	Yeast	cell-surface	display--applications	of	molecular	display.	Appl	
Microbiol	Biotechnol.	2004;64(1):28–40.		
	Korndorfer	IP,	Skerra	A.	Improved	affinity	of	engineered	streptavidin	for	the	Strep-tag	II	
peptide	is	due	to	a	fixed	open	conformation	of	the	lid-like	loop	at	the	binding	site.	
Protein	Sci.	2002;11:883–93.		
	
	
173	

	

Korosh	TC,	Dutcher	A,	Pfleger	BF,	McMahon	KD.	Inhibition	of	Cyanobacterial	Growth	on	a	
Municipal	Wastewater	Sidestream	Is	Impacted	by	Temperature.	mSphere.	
2018;3(1):1–15.		
	Koschwanez	JH,	Foster	KR,	Murray	AW.	Improved	use	of	a	public	good	selects	for	the	
evolution	of	undifferentiated	multicellularity.	Elife.	2013(2):e00367.	
	Koskella	B,	Brockhurst	MA.	Bacteria-phage	coevolution	as	a	driver	of	ecological	and	
evolutionary	processes	in	microbial	communities.	FEMS	Microbiol	Rev.	
2014;38(5):916–31.		
	Kumar	KH,	Jagadeesh	KS.	Microbial	consortia-mediated	plant	defense	against	
phytopathogens	and	growth	benefits.	South	Indian	J	Biol	Sci.	2016;2(4):395–403.		
	Kvennefors	ECE,	Sampayo	E,	Kerr	C,	Vieira	G,	Roff	G,	Barnes	AC,	et	al.	Regulation	of	
Bacterial	Communities	Through	Antimicrobial	Activity	by	the	Coral	Holobiont.	
Microbial	Ecology.	2012;63(3):605–18.		
	de	las	Heras	A,	Chavarría	M,	de	Lorenzo	V.	Association	of	dnt	genes	of	Burkholderia	sp.	DNT	
with	the	substrate-blind	regulator	DntR	draws	the	evolutionary	itinerary	of	2,4-
dinitrotoluene	biodegradation.	Mol	Microbiol.	2011;82(2):287–99.		
	Latifi	A,	Ruiz	M,	Zhang	CC.	Oxidative	stress	in	cyanobacteria.	FEMS	Microbiol	Rev.	
2009;33(2):258–78.		
	Latorre	C,	Lee	JH,	Spiller	H,	Shanmugam	KT.	Ammonium	ion-excreting	cyanobacterial	
mutant	as	a	source	of	nitrogen	for	growth	of	rice:	A	feasibility	study.	Biotechnol	Lett.	
1986;8(7):507–12.		
	Law	R,	Daniel	Morton	R.	Permanence	and	the	assembly	of	ecological	communities.	Ecology.	
1996;77(3):762–75.		
	LeClair	JS,	Wahl	LM.	The	Impact	of	Population	Bottlenecks	on	Microbial	Adaptation.	J	Stat	
Phys.	Springer	US;	2017;172(1):1–12.		
	Leino	H,	Kosourov	SN,	Saari	L,	Sivonen	K,	Tsygankov	AA,	Aro	EM,	et	al.	Extended	H2	
photoproduction	by	N2-fixing	cyanobacteria	immobilized	in	thin	alginate	films.	Int	J	
Hydrogen	Energy.	2012;37(1):151–61.	
	Lenski	RE.	Convergence	and	Divergence	in	a	Long-Term	Experiment	with	Bacteria.	Am	Nat.	
2017;190(S1):S57–68.	
	Lenski	RE,	Mongold	JA,	Sniegowski	PD,	Travisano	M,	Vasi	F,	Gerrish	PJ,	et	al.	Evolution	of	
competitive	fitness	in	experimental	populations	of	E.	coli:	What	makes	one	genotype	a	
better	competitor	than	another?	Antonie	van	Leeuwenhoek,	Int	J	Gen	Mol	Microbiol.	
1998;73(1):35–47.		
	

174	

	

	Lenski	RE,	Rose	MR,	Simpson	SC,	Tadler	SC.	Long-Term	Experimental	Evolution	in	
Escherichia	coli	.	I	.	Adaptation	and	Divergence	During	2	,	000	Generations.	Am	Nat.	
1991;138(6):1315–41.		
	Li	T,	Li	C-T,	Butler	K,	Hays	SG,	Guarnieri	MT,	Oyler	GA,	et	al.	Mimicking	lichens:	
Incorporation	of	yeast	strains	together	with	sucrose-secreting	cyanobacteria	improves	
survival,	growth,	ROS	removal,	and	lipid	production	in	a	stable	mutualistic	co-culture	
production	platform.	Biotechnol	Biofuels.	2017;10(1):1–12.		
	Liaimer	A,	Helfrich	EJN,	Hinrichs	K,	Guljamow	A,	Ishida	K,	Hertweck	C,	et	al.	Nostopeptolide	
plays	a	governing	role	during	cellular	differentiation	of	the	symbiotic	cyanobacterium	
Nostoc	punctiforme.	Proc	Natl	Acad	Sci.	2015;112(6):1862–7.		
	Lindemann	SR,	Bernstein	HC,	Song	HS,	Fredrickson	JK,	Fields	MW,	Shou	W,	et	al.	
Engineering	microbial	consortia	for	controllable	outputs.	ISME	J.	2016.	p.	2077–84.		
	Little	AE,	Robinson	CJ,	Peterson	SB,	Raffa	KF,	Handelsman	J.	Rules	of	Engagement:	
Interspecies	Interactions	that	Regulate	Microbial	Communities.	Annu	Rev	Microbiol.	
2008;62:375–401.	
	Liu	Z,	Müller	J,	Li	T,	Alvey	RM,	Vogl	K,	Frigaard	N,	et	al.	Genomic	analysis	reveals	key	
aspects	of	prokaryotic	symbiosis	in	the	phototrophic	consortium	“Chlorochromatium	
aggregatum”.	Genome	Biol.	2013;14(11):R127.	
	Löwe	H,	Hobmeier	K,	Moos	M,	Kremling	A,	Pflüger-Grau	K.	Photoautotrophic	production	of	
polyhydroxyalkanoates	in	a	synthetic	mixed	culture	of	Synechococcus	elongatus	cscB	
and	Pseudomonas	putida	cscAB.	Biotechnol	Biofuels.	2017a;:1–11.		
	Löwe	H,	Schmauder	L,	Hobmeier	K,	Kremling	A,	Pflüger-Grau	K.	Metabolic	engineering	to	
expand	the	substrate	spectrum	of	Pseudomonas	putida	toward	sucrose.	
Microbiologyopen.	2017b;6(4):1–9.		
	Löwe	H,	Sinner	P,	Kremling	A,	Pflüger-Grau	K.	Engineering	sucrose	metabolism	in	
Pseudomonas	putida	highlights	the	importance	of	porins.	Microb	Biotechnol.	
2018;0(0):1–10.		
	Lu	Y.	Making	Yeast	Competent	Cells	and	Yeast	Cell	Transformation.	Bio-Protocol.	
2011;1(14):4–6.		
	MacArthur	R,	Levins	R.	Competition,	habitat	selection,	and	character	displacement	in	a	
patchy	environment.	Proc	Natl	Acad	Sci.	1964;51(6):1207–10.		
	Maharjan	R,	Seeto	S,	Notley-McRobb	L,	Ferenci	T.	Clonal	Adaptive	Radiation	in	a	Constant	
Environment.	Science.	2006;313(5786):514–7.		
	
	
175	

	

Malmstrom	RR,	Coe	A,	Kettler	GC,	Martiny	AC,	Frias-Lopez	J,	Zinser	ER,	et	al.	Temporal	
dynamics	of	Prochlorococcus	ecotypes	in	the	Atlantic	and	pacific	oceans.	ISME	J.	
2010;4(10):1252–64.		
	Marchal	M,	Goldschmidt	F,	Derksen-Müller	SN,	Panke	S,	Ackermann	M,	Johnson	DR.	A	
passive	mutualistic	interaction	promotes	the	evolution	of	spatial	structure	within	
microbial	populations.	BMC	Evol	Biol.	2017;17(1):106.		
	Markson	JS,	Piechura	JR,	Puszynska	AM,	O’Shea	EK.	Circadian	control	of	global	gene	
expression	by	the	cyanobacterial	master	regulator	RpaA.	Cell.	2013;155(6):1396–408.		
	Martínez-García	E,	Jatsenko	T,	Kivisaar	M,	de	Lorenzo	V.	Freeing	Pseudomonas	putida	
KT2440	of	its	proviral	load	strengthens	endurance	to	environmental	stresses.	Environ	
Microbiol.	2015;17(1):76–90.		
	Mata	TM,	Martins	AAA,	Caetano	NS.	Microalgae	for	biodiesel	production	and	other	
applications:	A	review.	Renew	Sustain	Energy	Rev.	2010;14(1):217–32.	
	May	RM.	Thresholds	and	breakpoints	in	ecosystems	with	a	multiplicity	of	stable	states.	
Nature.	1977.	p.	471–7.		
	Mccann	KS.	The	diversity–stability	debate.	Nature.	2000;405:228–33.		
	McEwen	JT,	Machado	IMP,	Connor	MR,	Atsumi	S.	Engineering	Synechococcus	elongatus	PCC	
7942	for	continuous	growth	under	diurnal	conditions.	Appl	Environ	Microbiol.	2013	
Mar;79(5):1668–75.		
	McGrady-Steed	J,	Harris	PM,	Morin	PJ.	Biodiversity	regulates	ecosystem	predictability.	
Nature.	1997;390(6656):162–5.		
	Meeks	JC,	Elhai	J.	Regulation	of	Cellular	Differentiation	in	Filamentous	Cyanobacteria	in	
Free-Living	and	Plant-Associated	Symbiotic	Growth	States.	Microbiol	Mol	Biol	Rev.	
2002;66(1):94–121.		
	Messner	P,	Sleytr	UB.	Crystalline	bacterial	cell-surface	layers.	Adv	Microb	Physiol.	
1992;33:213–75.	
	Messner	P,	Steiner	K,	Zarschler	K,	Schäffer	C.	S-layer	nanoglycobiology	of	bacteria.	
Carbohydr	Res.	2008;343(12):1934–51.		
	Minty	JJ,	Singer	ME,	Scholz	SA,	Bae	C-H,	Ahn	J-H,	Foster	CE,	et	al.	Design	and	
characterization	of	synthetic	fungal-bacterial	consortia	for	direct	production	of	
isobutanol	from	cellulosic	biomass.	Proc	Natl	Acad	Sci	U	S	A.	2013;110(36):14592–7.	
	Momeni	B,	Chen	CC,	Hillesland	KL,	Waite	A,	Shou	W.	Using	artificial	systems	to	explore	the	
ecology	and	evolution	of	symbioses.	Cell.	Mol.	Life	Sci.	2011.	p.	1353–68.		
	

176	

	

	Momeni	B,	Waite	A,	Shou	W.	Spatial	self-organization	favors	heterotypic	cooperation	over	
cheating.	Elife.	2013;2013(2).		
	Morel	A,	Ahn	Y-H,	Partensky	F,	Vaulot	D,	Claustre	H.	Prochlorococcus	and	Synechococcus	-	
a	Comparative-Study	of	Their	Optical-Properties	in	Relation	To	Their	Size	and	
Pigmentation.	J	Mar	Res.	1993;51:617–49.		
	Moreno-Garrido	I.	Microalgae	immobilization :	Current	techniques	and	uses.	Bioresour	
Technol.	2008;99:3949–64.		
	Morris	JJ,	Johnson	ZI,	Szul	MJ,	Keller	M,	Zinser	ER.	Dependence	of	the	cyanobacterium	
Prochlorococcus	on	hydrogen	peroxide	scavenging	microbes	for	growth	at	the	ocean’s	
surface.	PLoS	One.	2011;6(2).		
	Moutinho	TJ,	Panagides	JC,	Biggs	MB,	Medlock	GL,	Kolling	GL,	Papin	JA.	Novel	co-culture	
plate	enables	growth	dynamic-based	assessment	of	contact-independent	microbial	
interactions.	PLoS	One.	2017;12(8):1–13.		
	Mueller	UG,	Sachs	JL.	Engineering	Microbiomes	to	Improve	Plant	and	Animal	Health.	
Trends	Microbiol.	2015.	p.	606–17.		
	Naeem	S,	Li	S.	Biodiversity	enhances	ecosystem	reliability.	Nature.	1997;390(6659):507–9.		
	Nai	C,	Meyer	V.	From	Axenic	to	Mixed	Cultures:	Technological	Advances	Accelerating	a	
Paradigm	Shift	in	Microbiology.	Trends	Microbiol.	2018.	p.	538–54.		
	Natrah	FMI,	Bossier	P,	Sorgeloos	P,	Yusoff	FM,	Defoirdt	T.	Significance	of	microalgal-
bacterial	interactions	for	aquaculture.	Rev.	Aquac.	2014.	p.	48–61.		
	Nelson	KE,	Weinel	C,	Paulsen	IT,	Dodson	RJ,	Hilbert	H,	Santos	VAPM,	et	al.	Complete	
genome	sequence	and	comparative	analysis	of	the	metabolically	versatile	
Pseudomonas	putida	KT2440.	Environ	Microbiol.	2002;4(12):799–808.		
	Niederholtmeyer	H,	Wolfstädter	BT,	Savage	DF,	Silver	P	a,	Way	JC.	Engineering	
cyanobacteria	to	synthesize	and	export	hydrophilic	products.	Appl	Environ	Microbiol.	
2010;76(11):3462–6.	
	Nikel	PI,	Chavarría	M,	Danchin	A,	de	Lorenzo	V.	From	dirt	to	industrial	applications:	
Pseudomonas	putida	as	a	Synthetic	Biology	chassis	for	hosting	harsh	biochemical	
reactions.	Curr	Opin	Chem	Biol.	2016;34:20–9.		
	Nikel	PI,	Chavarrı	M.	Endogenous	Stress	Caused	by	Faulty	Oxidation	Reactions	Fosters	
Evolution	of	2,4-Dinitrotoluene-	Degrading	Bacteria.	2013;9(8).		
	Nikel	PI,	Silva-Rocha	R,	Benedetti	I,	de	Lorenzo	V.	The	private	life	of	environmental	
	
177	

	

bacteria:	Pollutant	biodegradation	at	the	single	cell	level.	Environ	Microbiol.	
2014;16(3):628–42.		
	Nishino	SF,	Paoli	GC.	Aerobic	Degradation	of	Dinitrotoluenes	and	Pathway	for	Bacterial	
Degradation	of	Aerobic	Degradation	of	Dinitrotoluenes	and	Pathway	for	Bacterial	
Degradation	of	2,6-Dinitrotoluene.	Appl	Environ	Microbiol.	2000;66(5):2139–47.		
	Oh	S,	Seo	Y,	Ryu	K.	Reductive	removal	of	2,4-dinitrotoluene	and	2,4-dichlorophenol	with	
zero-valent	iron-included	biochar.	Bioresour	Technol.	2016;216:1014–21.	
	Ortiz-Marquez	JCF,	Do	Nascimento	M,	Curatti	L.	Metabolic	engineering	of	ammonium	
release	for	nitrogen-fixing	multispecies	microbial	cell-factories.	Metab	Eng.	
2014;23:154–64.		
	Ortiz-Marquez	JCF,	Do	Nascimento	M,	Dublan	MDLA,	Curatti	L.	Association	with	an	
ammonium-excreting	bacterium	allows	diazotrophic	culture	of	oil-rich	eukaryotic	
microalgae.	Appl	Environ	Microbiol.	2012;78(7):2345–52.	
	Ortiz-Marquez	JCF,	Do	Nascimento	M,	Zehr	JP,	Curatti	L.	Genetic	engineering	of	
multispecies	microbial	cell	factories	as	an	alternative	for	bioenergy	production.	
Trends	Biotechnol.	2013;31(9):521–9.	
	Padda	RS,	Wang	C,	Hughes	JB,	Kutty	R,	Bennett	GN.	Mutagenicity	of	nitroaromatic	
degradation	compounds.	Environ	Toxicol	Chem.	2003;22(10):2293–7.		
	Pande	S,	Kaftan	F,	Lang	S,	Svatoš	A,	Germerodt	S,	Kost	C.	Privatization	of	cooperative	
benefits	stabilizes	mutualistic	cross-feeding	interactions	in	spatially	structured	
environments.	ISME	J.	2016;10(6):1413–23.		
	Pandey	RP,	Parajuli	P,	Koffas	MAG,	Sohng	JK.	Microbial	production	of	natural	and	non-
natural	flavonoids:	Pathway	engineering,	directed	evolution	and	systems/synthetic	
biology.	Biotechnol	Adv.	2016;34(5):634–62.	
	Pankratov	TA,	Kachalkin	A	V,	Korchikov	ES,	Dobrovol	TG.	Microbial	Communities	of	
Lichens.	2017;86(3):293–309.		
	Papini	M,	Salazar	M,	Nielsen	J.	Systems	biology	of	industrial	microorganisms.	Adv	Biochem	
Eng	Biotechnol.	2010;120:51–99.	
	Park	W,	Peña-Llopis	S,	Lee	Y,	Demple	B.	Regulation	of	superoxide	stress	in	Pseudomonas	
putida	KT2440	is	different	from	the	SoxR	paradigm	in	Escherichia	coli.	Biochem	
Biophys	Res	Commun.	2006;341(1):51–6.		
	Park	Y,	Je	KW,	Lee	K,	Jung	SE,	Choi	TJ.	Growth	promotion	of	Chlorella	ellipsoidea	by	co-
inoculation	with	Brevundimonas	sp.	isolated	from	the	microalga.	Hydrobiologia.	
2008;598(1):219–28.		
	
178	

	

	Partensky	F,	Hess	WR,	Vaulot	D.	Prochlorococcus,	a	marine	photosynthetic	prokaryote	of	
global	significance.	Microbiol	Mol	Biol	Rev.	1999;63(1):106–27.	
	Paul	C,	Barofsky	A,	Vidoudez	C,	Pohnert	G.	Diatom	exudates	influence	metabolism	and	cell	
growth	of	co-cultured	diatom	species.	Mar	Ecol	Prog	Ser.	2009;389(2007):61–70.		
	Paul	C,	Mausz	MA,	Pohnert	G.	A	co-culturing/metabolomics	approach	to	investigate	
chemically	mediated	interactions	of	planktonic	organisms	reveals	influence	of	bacteria	
on	diatom	metabolism.	Metabolomics.	2013;9(2):349–59.		
	Pepper	L,	Cho	Y,	Boder	E,	Shusta	E.	A	decade	of	yeast	surface	display	technology:	where	are	
we	now?	Comb	Chem.	2008;11(2):127–34.	
	Pereira	SB,	Mota	R,	Vieira	CP,	Vieira	J,	Tamagnini	P.	Phylum-wide	analysis	of	
genes/proteins	related	to	the	last	steps	of	assembly	and	export	of	extracellular	
polymeric	substances	(EPS)	in	cyanobacteria.	Sci	Rep.	2015	6;5:14835.		
	Pérez-Pantoja	D,	Nikel	PI,	Chavarría	M,	de	Lorenzo	V.	Endogenous	Stress	Caused	by	Faulty	
Oxidation	Reactions	Fosters	Evolution	of	2,4-Dinitrotoluene-Degrading	Bacteria.	PLoS	
Genet.	2013;9(8).		
	Piechura	JR,	Amarnath	K,	Shea	EKO.	Natural	changes	in	light	interact	with	circadian	
regulation	at	promoters	to	control	gene	expression	in	cyanobacteria.	2017;1–37.		
	Ponce-Toledo	RI,	Deschamps	P,	López-García	P,	Zivanovic	Y,	Benzerara	K,	Moreira	D.	An	
Early-Branching	Freshwater	Cyanobacterium	at	the	Origin	of	Plastids.	Curr	Biol.	
2017;27(3):386–91.		
	Posewitz	MC.	Metabolism	and	Genetics	of	Algal	Hydrogen	Production.	Microalgal	Hydrog	
Prod.	2018.	p.	167–88.		
	Prosser	JI,	Bohannan	BJM,	Curtis	TP,	Ellis	RJ,	Firestone	MK,	Freckleton	RP,	et	al.	The	role	of	
ecological	theory	in	microbial	ecology.	Nat	Rev	Microbiol.	2007;5(5):384.		
	Rai		a.	N,	Söderbäck	E,	Bergman	B.	Cyanobacterium±plant	symbioses.	New	Phytol.	
2000;147(3):449–81.		
	Rai	AN,	Bergman	B,	Rasmussen	U.	Cyanobacteria	in	symbiosis.	Rai	AN,	Bergman	B,	
Rasmussen	U,	editors.	Kluwer	Academic	Pub.;	2002.		
	Ramos	JL,	Duque	E,	Maria-Trinidad	G,	Godoy	P.	Mechanisms	of	solvent	tolerance	in	gram-
negative	bacteria.	Annu	Rev	Microbiol.	2002;56:743–68.		
	Rastogi	RP,	Sinha	RP.	Biotechnological	and	industrial	significance	of	cyanobacterial	
secondary	metabolites.	Biotechnol	Adv.	2009;27(4):521–39.	
	

179	

	

	Ratcliff	WC,	Denison	RF,	Borrello	M,	Travisano	M.	Experimental	evolution	of	
multicellularity.	Proc	Natl	Acad	Sci.	2012;109(5):1595–600.	
	Raven	JA.	Why	Are	There	No	Picoplanktonic	O(2)	Evolvers	with	Volumes	Less-Than	10(-
19)	M(3).	J	Plankton	Res.	1994;16(5):565–80.		
	Reddington	SC,	Howarth	M.	Secrets	of	a	covalent	interaction	for	biomaterials	and	
biotechnology:	SpyTag	and	SpyCatcher.	Curr	Opin	Chem	Biol.	2015.	29:94–9.		
	Rittmann	BE.	Opportunities	for	renewable	bioenergy	using	microorganisms.	Biotechnol	
Bioeng.	2008;100(2):203–12.	
	Rocheleau	S,	Kuperman	RG,	Simini	M,	Hawari	J,	Checkai	RT,	Thiboutot	S,	et	al.	Toxicity	of	2	,	
4-dinitrotoluene	to	terrestrial	plants	in	natural	soils.	Sci	Total	Environ.	
2010;408:3193–9.		
	Rollié	S,	Mangold	M,	Sundmacher	K.	Designing	biological	systems:	Systems	Engineering	
meets	Synthetic	Biology.	Chem.	Eng.	Sci.	2012.	p.	1–29.		
	Romo	S,	Perez-Martinez	C.	The	Use	of	Immobilization	in	Alginate	Beads	for	Long-Term	
Storage	of	Pseudanabaena	galeata	(Cyanobacteria)	in	the	Laboratory.	J	Phycol.	
1997;(33):1073–6.		
	Rossetti	V,	Schirrmeister	BE,	Bernasconi	M	V,	Bagheri	HC.	The	evolutionary	path	to	
terminal	differentiation	and	division	of	labor	in	cyanobacteria.	J	Theor	Biol.	
2010;262(1):23–34.	
	Roy	KD,	Marzorati	M,	Van	den	Abbeele	P,	Van	de	Wiele	T,	Boon	N.	Synthetic	microbial	
ecosystems:	an	exciting	tool	to	understand	and	apply	microbial	communities.	Environ	
Microbiol.	2014;16(6):1472–81.		
	Rubin	BE,	Wetmore	KM,	Price	MN,	Diamond	S,	Shultzaberger	RK,	Lowe	LC,	et	al.	The	
essential	gene	set	of	a	photosynthetic	organism.	Proc	Natl	Acad	Sci	U	S	A.	
2015;112(48):E6634-43.	
	Ruiz-güereca	DA,	Sánchez-saavedra	MP.	Growth	and	phosphorus	removal	by	
Synechococcus	elongatus	co-immobilized	in	alginate	beads	with	Azospirillum	
brasilense.	J	Appl	Phycol.	Journal	of	Applied	Phycology;	2016;1501–7.		
	Sabra	W,	Dietz	D,	Tjahjasari	D,	Zeng	A-P.	Biosystems	analysis	and	engineering	of	microbial	
consortia	for	industrial	biotechnology.	Eng	Life	Sci.	2010;10(5):407–21.		
	Sabri	S,	Nielsen	LK,	Vickers	CE.	Molecular	control	of	sucrose	utilization	in	Escherichia	coli	
W,	an	efficient	sucrose-utilizing	strain.	Appl	Environ	Microbiol.	2013a;79(2):478–87.		
	
	

180	

	

Sabri	S,	Steen	JA,	Bongers	M,	Nielsen	LK,	Vickers	CE.	Knock-in/Knock-out	(KIKO)	vectors	
for	rapid	integration	of	large	DNA	sequences,	including	whole	metabolic	pathways,	
onto	the	Escherichia	coli	chromosome	at	well-characterised	loci.	Microb	Cell	Fact.	
2013b;12(1):1.	
	Said	S	Ben,	Or	D.	Synthetic	microbial	ecology:	Engineering	habitats	for	modular	consortia.	
Front	Microbiol.	2017;8(JUN).		
	Samuelson	P,	Gunneriusson	E,	Nygren	PA,	Ståhl	S.	Display	of	proteins	on	bacteria.	J	
Biotechnol.	2002;96(2):129–54.	
	Sandor	M,	Langone	J.	Demonstration	of	high	and	low	affinity	complexes	between	protein	A	
and	rabbit	immunoglobulin	C	antibodies	depending	on	hapten	density.	Biochem	
Biophys	Res.	1982;106(3):761–7.	
	Sauer	J,	Schreiber	U,	Schmid	R,	Völker	U,	Forchhammer	K.	Nitrogen	starvation-induced	
chlorosis	in	Synechococcus	PCC	7942.	Low-level	photosynthesis	as	a	mechanism	of	
long-term	survival.	Plant	Physiol.	2001;126(1):233–43.	
	Savage	VM,	Webb	CT,	Norberg	J.	A	general	multi-trait-based	framework	for	studying	the	
effects	of	biodiversity	on	ecosystem	functioning.	J	Theor	Biol.	2007;247(2):213–29.		
	Saxena	S.	Diversity	of	Industrially	Relevant	Microbes.	Appl	Microbiol.	New	Delhi:	Springer	
India;	2015.	p.	1–11.	
	Schmidt	TGM,	Skerra	A.	One-step	affinity	purification	of	bacterially	produced	proteins	by	
means	of	the	“Strep	tag”	and	immobilized	recombinant	core	streptavidin.	J	Chromatogr	
A.	1994;676(2):337–45.		
	Schuster	B,	Sleytr	UB.	S-layer-supported	lipid	membranes.	J	Biotechnol.	2000;74(3):233–
54.	
	Shaner	NC,	Lambert	GG,	Chammas	A,	Ni	Y,	Cranfill	PJ,	Baird	MA,	et	al.	A	bright	monomeric	
green	fluorescent	protein	derived	from	Branchiostoma	lanceolatum.	Nat	Methods.	
2013;10(5):407–9.		
	Sharma	NK,	Tiwari	SP,	Tripathi	K,	Rai	AK.	Sustainability	and	cyanobacteria	(blue-green	
algae):	Facts	and	challenges.	J	Appl	Phycol.	2011;23(6):1059–81.		
	Sheehan	J,	Dunahay	T,	Benemann	J,	Roessler	P.	Look	Back	at	the	U.S.	Department	of	
Energy’s	Aquatic	Species	Program:	Biodiesel	from	Algae;	Close-Out	Report.	
1998;(July).		
	Shemer	B,	Yagur-kroll	S,	Hazan	C,	Belkin	S.	Aerobic	Transformation	of	2,4-Dinitrotoluene	
by	Escherichia	coli	and	Its	Implications	for	the	Detection	of	Trace	Explosives.	Appl	
Environ	Microbiol.	2018;84(4):e01729-17.		
	
181	

	

	Sher	D,	Thompson	JW,	Kashtan	N,	Croal	L,	Chisholm	SW.	Response	of	Prochlorococcus	
ecotypes	to	co-culture	with	diverse	marine	bacteria.	ISME	J.	2011.	p.	1125–32.		
	Sherr	EB,	Sherr	BF.	Significance	of	predation	by	protists	in	aquatic	microbial	food	webs.	
Antonie	van	Leeuwenhoek,	Int	J	Gen	Mol	Microbiol.	2002;81(1–4):293–308.		
	Shevchenko	A,	Wilm	M,	Vorm	O,	Mann	M.	Mass	Spectrometric	Sequencing	of	Proteins	from	
Silver-Stained	Polyacrylamide	Gels.	Anal	Chem.	1996;68(5):850–8.		
	Shiloach	J,	Bauer	S,	Chemistry	B.	High-Yield	Growth	of	E.	coli	at	Different	Temperatures	in	a	
Bench	Scale	Fermentor.	Biotechnol	Bioeng.	1975;XVII:227–39.		
	Shong	J,	Jimenez	Diaz	MR,	Collins	CH.	Towards	synthetic	microbial	consortia	for	
bioprocessing.	Curr	Opin	Biotechnol.	2012;23(5):798–802.	
	Shou	W,	Ram	S,	Vilar	JMG.	Synthetic	cooperation	in	engineered	yeast	populations.	Proc	Natl	
Acad	Sci	USA.	2007;104(6):1877–82.	
	Simkovsky	R,	Daniels	EF,	Tang	K,	Huynh	SC,	Golden	SS,	Brahamsha	B.	Impairment	of	O-
antigen	production	confers	resistance	to	grazing	in	a	model	amoeba-cyanobacterium	
predator-prey	system.	Proc	Natl	Acad	Sci.	2012;109(41):16678–83.		
	Singh	AP,	Tiwari	DN.	Phenotypic	expression	of	ammonia-excreting	mutants	of	Anabaena	
7120	under	nitrogen	limitation.	World	J	Microbiol	Biotechnol.	1998;14(4):591–3.		
	Šmarda	J,	Šmajs	D,	Komrska	J,	Krzyžánek	V.	S-layers	on	cell	walls	of	cyanobacteria.	Micron.	
2002;33(3):257–77.	
	Smith	MJ,	Francis	MB.	A	Designed	A.	vinelandii−S.	elongatus	Coculture	for	Chemical	
Photoproduction	from	Air,	Water,	Phosphate,	and	Trace	Metals.	ACS	Synth	Biol.	
2016;5:acssynbio.6b00107.	
	Smith	MJ,	Francis	MB.	Improving	metabolite	production	in	microbial	co-cultures	using	a	
spatially	constrained	hydrogel.	Biotechnol	Bioeng.	2017;114(6):1195–200.		
	Smith	VH,	Sturm	BSM,	deNoyelles	FJ,	Billings	SA.	The	ecology	of	algal	biodiesel	production.	
Trends	Ecol.	Evol.	2010.	p.	301–9.		
	Söding	J,	Biegert	A,	Lupas	AN.	The	HHpred	interactive	server	for	protein	homology	
detection	and	structure	prediction.	Nucleic	Acids	Res.	2005;33(SUPPL.	2):244–8.		
	Song	H-S,	Cannon	W,	Beliaev	A,	Konopka	A.	Mathematical	Modeling	of	Microbial	
Community	Dynamics:	A	Methodological	Review.	Processes.	2014a;2(4):711–52.		
	Song	H,	Ding	M-Z,	Jia	X-Q,	Ma	Q,	Yuan	Y-J.	Synthetic	microbial	consortia:	from	systematic	
	

182	

	

analysis	to	construction	and	applications.	Chem	Soc	Rev.	2014b;	
	Song	K,	Tan	X,	Liang	Y,	Lu	X.	The	potential	of	Synechococcus	elongatus	UTEX	2973	for	sugar	
feedstock	production.	Appl	Microbiol	Biotechnol.	2016;(189).	
	Spain	JC.	Biodegradation	of	Nitroaromatic	Compounds.	Annu	Rev	Microbiol.	1995;49:2548-
2523–55.		
	Spanggord	RJ,	Spain	JC,	Nishino	SF,	Mortelmans	KE.	Biodegradation	of	2,4-dinitrotoluene	
by	a	Pseudomonas	sp.	Appl	Environ	Microbiol.	1991;57(11):3200–5.		
	Spiller	H,	Gunasekaran	M.	Ammonia-excreting	mutant	strain	of	the	cyanobacterium	
Anabaena	variabilis	supports	growth	of	wheat.	Appl	Microbiol	Biotechnol.	
1990;33(4):477–80.		
	Spiller	H,	Latorre	C,	Hassan	ME,	Shanmugam	AKT.	Isolation	and	Characterization	of	
Nitrogenase-Derepressed	Mutant	Strains	of	Cyanobacterium	Anabaena	variabilist.	J	
Bacteriol.	1986;165(2):412–9.		
	Srinivasulu	B,	Adinarayana	K,	Ellaiah	P.	Investigations	on	Neomycin	Production	With	
Immobilized	Cells	of	Strep-	tomyces	marinensis	Nuv-5	in	Calcium	Alginate	Matrix.	
2003;4(4):1–6.		
	Stapleton	J	a.,	Whitehead	T	a.,	Nanda	V.	Computational	redesign	of	the	lipid-facing	surface	
of	the	outer	membrane	protein	OmpA.	Proc	Natl	Acad	Sci.	2015;112(31):9632–7.	
	Stoodley	P,	Sauer	K,	Davies	DG,	Costerton	JW.	Biofilms	as	complex	differentiated	
communities.	Annu	Rev	Microbiol.	2002;56:187–209.	
	Sumper	M,	Berg	E,	Mengele	R.	Primary	structure	and	glycosylation	of	the	S-layer	protein	of	
Haloferax	volcanii.	J	Bacteriol.	1990;172(12):7111–8.		
	Symons	ZC,	Bruce	NC.	Bacterial	pathways	for	degradation	of	nitroaromatics.	Nat	Prod	Rep.	
2006;23(6):845.	
	Tan	J,	Zuniga	C,	Zengler	K.	Unraveling	interactions	in	microbial	communities	-	from	co-
cultures	to	microbiomes.	J	Microbiol.	2015;53(5):295–305.		
	Tandon	P,	Jin	Q.	Microalgae	culture	enhancement	through	key	microbial	approaches.	
Renew.	Sustain.	Energy	Rev.	2017.	p.	1089–99.		
	Tecon	R,	Or	D.	Cooperation	in	carbon	source	degradation	shapes	spatial	self-	organization	
of	microbial	consortia	on	hydrated	surfaces.	Sci	Rep.	2017;7:1–11.	
	Temperton	B,	Giovannoni	SJ.	Metagenomics:	Microbial	diversity	through	a	scratched	lens.	
Curr.	Opin.	Microbiol.	2012.	p.	605–12.		
	
183	

	

	Therien	JB,	Zadvornyy	OA,	Posewitz	MC,	Bryant	DA,	Peters	JW.	Growth	of	Chlamydomonas	
reinhardtii	in	acetate-free	medium	when	co-cultured	with.	2014;1–8.		
	Tizei	PAG,	Csibra	E,	Torres	L,	Pinheiro	VB.	Selection	platforms	for	directed	evolution	in	
synthetic	biology.	Biochem	Soc	Trans.	2016;44(4):1165–75.	
	Tong	Z,	Zhe	L,	Huai-lan	Z.	microbial	flocculant	and	its	application	in	environmental	
protection.	J	Env	Sci.	1999;11(1):1–12.	
	Trautner	C,	Vermaas	WFJ.	The	sll1951	gene	encodes	the	surface	layer	protein	of	
Synechocystis	sp.	strain	PCC	6803.	J	Bacteriol.	2013;195(23):5370–80.	
	Tringe	SG,	Rubin	EM.	Metagenomics:	DNA	sequencing	of	environmental	samples.	Nat.	Rev.	
Genet.	2005.	p.	805–14.		
	Trivedi	P,	Delgado-Baquerizo	M,	Trivedi	C,	Hu	H,	Anderson	IC,	Jeffries	TC,	et	al.	Microbial	
regulation	of	the	soil	carbon	cycle:	Evidence	from	gene-enzyme	relationships.	ISME	J.	
2016;10(11):2593–604.	
	Tsimring	L,	Hasty	J,	Danino	T,	Mondrago	O,	Mondragón-Palomino	O,	Tsimring	L,	et	al.	A	
synchronized	quorum	of	genetic	clocks.	Nature.	2010;463(7279):326–30.		
	Tsutsui	H,	Karasawa	S,	Okamura	Y,	Miyawaki	A.	Improving	membrane	voltage	
measurements	using	FRET	with	new	fluorescent	proteins.	Nat	Methods.	
2008;5(8):683–5.		
	Umeda	H,	Aiba	H,	Mizuno	T.	SomA,	a	novel	gene	that	encodes	a	major	outer-membrane	
protein	of	Synechococcus	sp.	PCC	7942.	Microbiology.	1996;142	(	Pt	8:2121–8.		
	Usher	KM,	Bergman	B,	Raven	JA.	Exploring	Cyanobacterial	Mutualisms.	Annu	Rev	Ecol	Evol	
Syst.	2007;38(1):255–73.		
	Vaara	T.	The	outermost	surface	structures	in	chroococcacean	cyanobacteria.	Can	J	
Microbiol.	1982;28:929–41.	
	Vaishampayan	A,	Sinha	RP,	Häder	DP,	Dey	T,	Gupta	AK,	Bhan	U,	et	al.	Cyanobacterial	
biofertilizers	in	rice	agriculture.	Bot.	Rev.	2001.	p.	453–516.		
	Verseux	CN,	Baqué	M,	Lehto	K,	De	Vera	JPP,	Rothschild	LJ,	Billi	D.	Sustainable	life	support	
on	Mars	-	The	potential	roles	of	cyanobacteria.	Int	J	Astrobiol.	2016.	p.	65–92.		
	Verseux	CN,	Paulino-Lima	IG,	Baqué	M,	Billi	D,	Rothschild	LJ.	Synthetic	biology	for	space	
exploration:	Promises	and	societal	implications.	Ambivalences	Creat	Life	Soc	Philos	
Dimens	Synth	Biol.	2015.	p.	73–100.		
	
	

184	

	
Voss	S,	Skerra		a.	Mutagenesis	of	a	flexible	loop	in	streptavidin	leads	to	higher	affinity	for	
the	Strep-tag	II	peptide	and	improved	performance	in	recombinant	protein	
purification.	Protein	Eng.	1997;10(8):975–82.		
	Wahl	LM,	Gerrish	PJ,	Saika-Voivod	I.	Evaluating	the	impact	of	population	bottlenecks	in	
experimental	evolution.	Genetics.	2002;162(2):961–71.		
	Waite	AJ,	Shou	W.	Adaptation	to	a	new	environment	allows	cooperators	to	purge	cheaters	
stochastically.	Proc	Natl	Acad	Sci.	2012;109(47):19079–86.		
	Wang	H,	Zhang	W,	Chen	L,	Wang	J,	Liu	T.	The	contamination	and	control	of	biological	
pollutants	in	mass	cultivation	of	microalgae.	Bioresour.	Technol.	2013.	p.	745–50.		
	Wang	Z,	Ye	Z,	Zhang	M.	Bioremediation	of	2,4-dinitrotoluene	(2,4-DNT)	in	immobilized	
micro-organism	biological	filter.	J	Appl	Microbiol.	2011;(110):1476–84.		
	Waters	CM,	Bassler	BL.	Quorum	Sensing :	Communication	in	Bacteria.	Annu	Rev	Cell	Dev	
Biol.	2005;21(1):319–46.	
	Weiss	TL,	Ducat	DC.	Designing	stable,	synthetic,	light-driven	cyanobacteria-heterotroph	
consortia	for	bioproduction.	Metab	Eng.	2017;(44):236–45.		
	Weiss	TL,	Young	EJ,	Ducat	DC.	A	synthetic	,	light-driven	consortium	of	cyanobacteria	and	
heterotrophic	bacteria	enables	stable	polyhydroxybutyrate	production.	Metab	Eng.	
2017;44:236–45.		
	Werner	GDA,	Strassmann	JE,	Ivens	ABF,	Engelmoer	DJP,	Verbruggen	E,	Queller	DC,	et	al.	
Evolution	of	microbial	markets.	Proc	Natl	Acad	Sci	USA.	2014	Jan	28;111(4):1237–44.	
	Widder	S,	Allen	RJ,	Pfeiffer	T,	Curtis	TP,	Wiuf	C,	Sloan	WT,	et	al.	Challenges	in	microbial	
ecology:	Building	predictive	understanding	of	community	function	and	dynamics.	
ISME	J.	2016.	p.	2557–68.		
	Wilkinson	TG,	Topiwala	HH,	Hamer	G.	Interactions	in	a	mixed	bacterial	population	growing	
on	methane	in	continuous	culture.	Biotechnol	Bioeng.	1974;16(1):41–59.		
	Wintermute	EH,	Silver	PA.	Dynamics	in	the	mixed	microbial	concourse.	Genes	Dev.	
2010;24(23):2603–14.		
	Wyman	CE,	Goodman	BJ.	Biotechnology	for	production	of	fuels,	chemicals,	and	materials	
from	biomass.	Appl	Biochem	Biotechnol.	1993	Sep;39–40(1):41–59.	
	Xu	Y,	Tiago	Guerra	L,	Li	Z,	Ludwig	M,	Charles	Dismukes	G,	Bryant	D	a.	Altered	carbohydrate	
metabolism	in	glycogen	synthase	mutants	of	Synechococcus	sp.	strain	PCC	7002:	Cell	
factories	for	soluble	sugars.	Metab	Eng.	2012;16C:56–67.		
	
	

185	

	

Yang	S,	Sleight	SC,	Sauro	HM.	Rationally	designed	bidirectional	promoter	improves	the	
evolutionary	stability	of	synthetic	genetic	circuits.	Nucleic	Acids	Res.	2013;41(1):e33.	
	Ye	C,	Zou	W,	Xu	N,	Liu	L.	Metabolic	model	reconstruction	and	analysis	of	an	artificial	
microbial	ecosystem	for	vitamin	C	production.	J	Biotechnol.	2014;182–183(1):61–7.	
	Yinon	J.	Toxicity	and	Metabolism	of	Explosives.	Boca	Raton,	Florida,	USA:	CRC	Press,	Inc.;	
1990.		
	Yoon	JM,	Oliver	DJ,	Shanks	J	V.	Phytotransformation	of	2,4-Dinitrotoluene	in	Arabidopsis	
thaliana :	Toxicity	,	Fate	,	and	Gene	Expression	Studies	in	Vitro.	2006;1524–31.		
	Zakeri	B,	Fierer	JO,	Celik	E,	Chittock	EC,	Schwarz-Linek	U,	Moy	VT,	et	al.	Peptide	tag	forming	
a	rapid	covalent	bond	to	a	protein,	through	engineering	a	bacterial	adhesin.	Proc	Natl	
Acad	Sci	USA.	2012;109(12):E690-7.	
	Zavřel	T,	Sinetova	MA,	Červený	J.	Measurement	of	Chlorophyll	a	and	Carotenoids	
Concentration	in	Cyanobacteria.	Bio-Protocol.	2015;5(9):1–5.		
	Zengler	K,	Palsson	BO.	A	road	map	for	the	development	of	community	systems	(CoSy)	
biology.	Nat	Rev	Microbiol.	2012;10(5):366–72.	
	Zhang	CC,	Laurent	S,	Sakr	S,	Peng	L,	Bédu	S.	Heterocyst	differentiation	and	pattern	
formation	in	cyanobacteria:	A	chorus	of	signals.	Mol.	Microbiol.	2006.	p.	367–75.		
	Zhang	Z,	Meng	L,	Ni	C,	Yao	L,	Zhang	F,	Jin	Y,	et	al.	Engineering	Escherichia	coli	to	bind	to	
cyanobacteria.	J	Biosci	Bioeng.	2017;123(3):347–52.		
	Zinser	ER.	Cross-protection	from	hydrogen	peroxide	by	helper	microbes:	The	impacts	on	
the	cyanobacterium	Prochlorococcus	and	other	beneficiaries	in	marine	communities.	
Environ	Microbiol	Rep.	2018;00:1–13.		
	

	

186