'I‘ «shim s1? yew—M ‘ u .lg.n> -.u.. .3 .3 .7 .. «:.g:n. in- “g” 3.2"“"1 . .u -.rXJ L...,'. . . ‘3be ,4“ 3’2" ,.. .. 1 (J 1-» . :J.)'., V .r ,J'U‘ a J ;. N ,1; u. .1 I ‘ . . .A. - —_-4jo-P\¢-\-¢y< x. uftw tv 'JEJJR 11' UL L’J","‘L. ‘:J€:rx:’: x: mg "Mi. ‘ll 1“"? .. flab? V . ’5'). 52"“ £332: rig»; -‘ o“éfln.‘-' 0 V‘. ; WA l'afiz' .~ VVv'Hry fih‘éfi _,gza.z.-: x: ’7‘ TWA“ I- Iv. fl; ’film' W - . r-~»_ ‘ I. $._ l:- ‘0"; .‘l _ r. I [3:15; 5—2:: 3.; A i‘:¢v17 ,vfé‘k‘: 72::5 ' e a. "-31" ”,:‘| 9“." u m :r:;: :..:,;' <. C ..,_ <. r ‘r‘ .sm' ’Jb_ ‘v..~- I. L' “145 "u. i” :3“ u 7%. $ "2}“ 7.33. :X'Lt'“' a- x 27 4- 3‘ v: .3. n3. ,. , .. my). . ' ‘ 'N‘f; ‘ $3.1,” I fake“ "75271;“.3 ‘w' ‘ mu 'k- 3 ‘H: -"- | V ’12 ES H lnllfl lljflljlilllllllllllilll 00901 7058 This is to certify that the thesis entitled C SIGNAL-DEPENDENT, DEVELOPMENTAL GENE EXPRESSION IN MYXOFCOC'CVUS ‘XANEHES presented by Monica Ellen Semancik has been accepted towards fulfillment of the requirements for Master Science degree in 5m, 7mm Major professor Date 6/ 16/q ’ 0.7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State ’ University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before ode due. DATE DUE DATE DUE DATE DUE :ll :lll MSU Is An Atfinnetive Action/Equal Opportunity Institution owns-oi C SIGNAL-DEPENDENT, DEVELOPMENTAL GENE EXPRESSION IN MYXOCOCCUS XANTHUS by Monica Ellen Semancik A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Biochemistry 1991 ABSTRACT C SIGNAL-DEPENDENT, DEVELOPMENTAL GENE EXPRESSION IN MYXOCOCCUS XANTHUS by Monica Ellen Semancik Four classes of cell-cell interactions (A, B, C, and D) are absolutely required for starvation-induced morphogenesis and developmental gene expression in MYxococcus xanthus. To investigate the mechanisms which couple intercellular signaling and gene activation, DNA adjacent to previously identified C signal-dependent Tn51ac fusions was cloned and tested for the ability to direct developmental gene expression. A 2 kbp region adjacent to Tn51ac Q4403 promotes developmental B-galactosidase expression and at least partially' mediates ‘the C signal dependence. DNA between 2 and 8.5 kbp of the 94403 fusion also appears to positively affect the level of expression. Accumulation of the Q4403-associated transcript is developmentally regulated suggesting that control of transcription initiation may be one mechanism for modulating gene expression. To facilitate the identification of factors involved in C-dependent transcription, core RNA polymerase was purified from vegetative M. xanthus. Core RNA polymerase transcriptional activity is stimulated by both E. coli 0'70 and a 50 kDa M. xanthus protein. To my parents, John and Martha, whose love, patience, and unending encouragement are the cornerstones of my success and to Mark who has taught me that nothing is impossible (except skiing through a revolving door)! iii ACKNOWLEDGMENTS I would like to specially thank Lee Kroos for his guidance and support during the past two years. I am extremely grateful for the flexibility and independence he gave me in my research, the kindness he showed when I made mistakes, and the enthusiasm he shared when experiments succeeded! my deepest respect goes to him for encouraging excellence simply by his example. I never would have safely survived the many "lab emergencies" and "panic situations" without the quick- thinking abilities of Rich Halberg. For a mere "fifty bucks", Rich would load a protein gel, salvage a DNA- cellulose column, and give you the coat off his back! I thank Rich, as well as Sijie Lu, and co-Myxo worker Jamie White for their special friendship. My appreciation also goes to the members of my guidance committee, Zach Burton and Mike Tomashow, for their time, interest, and suggestions. I would like to acknowledge people who contributed materials to this research: R. Doi for the anti-Bacillus 043 antibodies, C. Gross for E. coli core RNA polymerase, G. Zeikus for E. coli strain JM83, and M. Sinton and D. iv Cryderman (in A. Revzin's lab) for E. coli holoenzyme, the lacUVS promoter, and use of their computer. Andrea Von‘rom deserves special thanks for critically reading my thesis, her helpful comments throughout the year, and most importantly for her friendship and encouragement (especially throughout these last "stressed-out" weeks!). Lastly, I wish to thank my fiance, Mark. His love and laughter have helped me enjoy the true "awesome-ness" of life. I also appreciate the excellent computer-graphics he generated for this thesis. TABLE OF CONTENTS Page LIST OF TABLES...0.0...OOOOOOOOOOOOOOOOOOOO0.... ix LIST OF FIGURES.......... ........... .... ...... .. x LIST OF ABBREVIATIONS....... .................... xii INTRODUCTIONOOOOO0....O0.0...OOOOOOOOOOOOOOOOOOO 1 CHAPTER 1: CHARACTERIZATION OF C-DEPENDENT REGULATORY REGIONS IN MYXOCOCCUS XANTHUS........ 10 INTRODUCTION......... ........................ 10 MATERIALS AND METHODS. O O. O O O O O O O O O O O O O O O O O O O O O 14 Bacterial strains and p1asmids.............. 14 Growth conditions..OOOOOOCOOCOOOOOO0..I0.... 14 Cloning strategy and plasmid construction... 14 Transfer of plasmid DNA from E. coli to M. xanthusoOOOOOO0.00000000000000000000000 22 Southern analysis........................... 23 B—galactosidase assays...................... 23 RNA iSOIationOOOOOOOOOOOOOOOOOOOOOOOOOOCO... 24 Sl-nuclease protection experiments.......... 25 RESULTS...O0....0......OOOCOOOOOOOOOOOCOOOOOO. 27 A 1.8 kbp Q4414 upstream region is insufficient for proper developmental gene expression................................ 28 Developmental gene expression can be directed by DNA within 2 kbp of the 04403 insertion site............................ 33 94403 is located approximately 380 bp downstream from the start of the transcription unit into which it has inserted............................. ..... 39 vi TABLE OF CONTENTS (con'd) Regulation of Q4403-associated gene expression occurs at the level of mRNA accumulationOOOOOO0.0.0.0....0.00....O...O C-dependent activation of gene expression occurs at least partially through DNA within 2 kbp of the 94403 fusion. . . . . . . . . . DISCUSSION ....................... . ............ CHAPTER 2: CHARACTERIZATION OF VEGETATIVE RNA POLYMERASE FROM MYXOCOCCUS XANTHUS......... ..... INTRODUCTION.COOOOOOCOOOOOOOOOOOOO0.... ........ MATERIALS AND METHODS........... .............. Growth of strains............... ............ Buffers........................ ............. RNA polymerase purification...... ........... In vitro transcription assays............... SDS-polyacrylamide gels.......... ....... .... Western blot analysis....................... RESULTS.IO....00...0......OOOOOOOOOOOOOOOOOOOO Purification of vegetative RNA polymerase by DNA-cellulose chromatography............. M. xanthus core polymerase activity can be stimulated by E. coli 070................ M. xanthus holoenzyme can utilize different bacterial promoters...................... A 50 kDa protein present in M. xanthus holoenzyme containing fractions is capable of stimulating transcription....... ...... DISCUSSIONOOOOOOOOOOO ..... O... OOOOOOOOOOOOOOO vii Page 43 44 51 57 57 60 60 6O 61 63 64 65 66 66 74 76 78 84 TABLE OF CONTENTS (con'd) Page SUMMARY AND CONCLUSIONS........................ 90 LIST OF REFERENCES.................. ...... ..... 94 viii LIST OF TABLES Table Page CHAPTER 1 1 Bacterial strains and plasmids ............. 15 CHAPTER 2 1 Comparison of several bacterial promoters.. 7? ix Figure 10 LIST OF FIGURES CHAPTER 1 Dependence of gene expression on the A, B, andcSignaISOOOOOOOOOOOOOOOOOOOOOOOOOOO... Cloning strategy for isolating DNA adjacent to developmentally-regulated Tn51ac fusions ...... ......... ........ ..... . ...... Chromosomal structures resulting from homologous and site-specific integration eventSOOOOOOOOOOOOOOOOOOOCOOOCOOO...0...... Developmental B-galactosidase expression from lacZ fused to the putative regulatory region upstream Of 04414.0.000000000000000 Developmental B-galactosidase expression from lacz fused to the putative regulatory region upstream of Q4403.................. Strategy for subcloning regulatory regions adjacent to “4403‘000000000000000000.00...0O Developmental B-galactosidase expression from lacz fused to portions of the 8.5 kbp regulatory region upstream of 04403 . . . . . . . . Developmental B-galactosidase expression in homologous integrants...................... Low resolution, S1-nuclease mapping of developmental transcripts from DK4368...... Quantitative Sl-nuclease protection of RNA from DK4368 throughout development......... Page 11 20 30 32 34 36 37 4O 42 46 Figure 11 LIST OF FIGURES (con'd) Developmental B-galactosidase expression from lacZ fused to the 94403 regulatory region in csgA strains..................... CHAPTER 2 SDS-polyacrylamide gel analysis of proteins in various fractions of the purification... DNA-cellulose chromatography........ ....... Run-off transcription by vegetative RNA polymerase from M. xanthus and B. subtilis. Stimulation of E. coli and M. xanthus core RNA polymerase transcriptional activity... Comparison of proteins present in fractions containing M. xanthus core and holoenzyme. Western blot analysis of core and holoenzyme-containing fractions........... Isolation of an M. xanthus transcription stimulating protein...... ...... ...... ..... xi Page 48 68 7O 72 75 79 81 83 DEPC DTT EDTA IgG KIU ONP ONPG PMSF PVDF SDS TAE TBS TCA Tris-HCl UAS LIST OF ABBREVIATIONS diethylpyrocarbonate dithiothreitol (ethelyenedinitrilo)tetraacetic acid immunoglobulin G Kallikrein-inhibitor units Klett units o-nitrophenol 2-nitrophenyl-b-D-galactopyranoside phenylmethylsulfonyl fluoride polyvinylidene difluoride sodium dodecyl sulfate Tris-acetate-EDTA buffer Tris buffered saline trichloroacetic acid 2-amino-2-(hydroxymethyl)-1,3-propandiol- hydrochloride upstream activation site xii INTRODUCTION Myxococcus xanthus is a Gram-negative prokaryote which offers a unique opportunity to study phenomena most often associated with higher multicellular organisms, particularly cell-cell interactions. Commonly found in the soil, groups of MYxococcus cells coordinately prey on other bacteria or feed on decomposing organic material by secreting a multitude of enzymes that degrade protein, lipid, peptidoglycan, and polysaccharide. In the laboratory, M. xanthus can be grown in liquid cultures or on solid agar containing hydrolyzed casein and a few salts (1). Organized groups of‘ Myxococcus tend to move together in "swarms" within which cell-cell contact is maintained, although independent cells are capable of a slow, gliding movement. Perhaps most interesting to developmental biologists are the morphogenesis and social interactions which occur in response to nutrient deprivation (2). Under conditions of amino acid starvation, approximately 105 individual cells participate in forming a three-dimensional fruiting body in which myxospores mature. Along with nutrient limiting conditions, a high initial cell density and a- solid surface are required for development (3,4). Development proceeds through a specific sequence of events (5). Within 5 to 7 hours after the onset of starvation, cells begin to accumulate in asymmetric ridges. 1 Shortly thereafter as cells continue to pile on top of one another, stable, circular mounds held together by an. extracellular matrix form at specific aggregation centers. As the fruiting bodies mature, cells within the structures begin 'to Idifferentiate from. long, rod-shaped cells into (ovoid myxospores. Several coat proteins are added to the exterior of the outer membrane and provide the spore with resistance to a variety of environmental stresses including heat, radiation, detergent, and dessication (6). Only about 20% of the cells actually complete the sporulation process (3) - The differentiation of myxobacteria is unique among the sporulation processes of other bacteria in that it absolutely requires cell-cell interactions. Four classes of conditional, non-sporulating mutants (A, B, C, D) have been isolated which arrest development at particular stages. The ability to sporulate can be recovered by mixing a mutant strain from one class with either wild-type cells or mutant cells from another class, suggesting these mutants are defective in afiouflu uca-~& Nona mewfimua area fill ......h.m.u.”....n.n..... m x Z \ / l LmJOEOLQ ELMO NUOH m>«JOJ3Q VA XJ. LDDOEOLQ m>«AOC cosmetmmncq TDOOOHOEOI /\ 31 uc.-_a LmDOEOLQ m>~A0JJQ \( Q ELmJ . mmwmmmua \\xe Emma Vxfl flwww .1 _ Tl Tilisx Il-xnuuuwu meuoxaneo aneoxmeuo Illlllmunnwlllll. . mXZ UC~I~Q LEI NUOH mxz mesa COLDOLOmDCL Usuaumam-musm m 32 4000 B-galactosldase specific activity (nmol ONP/mg protein'mm) 0 10 20 30 40 hours of development Figure 4. Developmental B-galactosidase expression from lacz fused to the putative regulatory region upstream of 04414. A promoterless lacz gene was fused to 1.8 kbp of DNA located immediately upstream from the (24414 insertion (pMESOOZ) and introduced into wild type M. xanthus at the Mx8 phage attachment site. B—galactosidase expression was measured in cells harvested at various times during their development on TPM starvation agar as described in the Materials and Methods. The specific activity of the wild type strain (DK1622; O—O), the original Q4414 fusion strain (DK5279; Cl—CI), and the average specific activity of two independently-isolated transductants containing a single copy of pMESOOZ integrated at the Mx8 attB site ( I—I). Error bars showing one standard deviation are too small to be seen. 33 isolated transductants containing a single copy of pMESOOZ integrated at the Mx8 attB site Clearly resembles the pattern seen in DK1622; only a basal level of expression is observed. This result indicates that 1.8 kbp of DNA upstream of (24414 is insufficient to direct developmental gene expression and implies that the promoter and possibly important regulatory regions lie further upstream. Dayalppmental gene sxprsssion can pe direpped by DNA wiphin 2 k of h (244 ins r ion sit . An 8.5 kbp segment of DNA adjacent to the 94403 insertion was cloned (pMESOO4) and reintroduced into wild- type M. xanthus. As illustrated in Figure 5, B- galactosidase activity starts to increase between 6 and 12 hours after the onset of starvation in the original Q4403 fusion strain (DK4368). Similarly, the average B- galactosidase activity of :3 independently-isolated transductants containing a single copy of pMESOO4 integrated at the Mx8 attB site increases during the same time interval, although the maximum level of expression for these ‘transductants is only' 25% of that observed for DK4368. These results suggest that at least some of the regulatory regions required for developmental gene expression are located within 8.5 kbp of the 94403 insertion point. In order to more closely define the regions within the 8.5 kbp of upstream DNA responsible for directing B- galactosidase expression, portions of this segment were 34 3m) B-galactosidase specific activity (nmol ONP/mg protem'min) 0 20 40 60 80 hours of development Figure 5. Developmental B-galactosidase expression from lacz fused to the putative regulatory region upstream of 04403. A promoterless lacz gene was fused to 8.5 kbp of DNA located immediately upstream from the 94403 insertion (pMESOO4) and introduced into wild type M. xanthus at the Mx8 phage attachment site. B-galactosidase expression was measured in cells harvested at various times during their development on TPM starvation agar as described in the Materials and Methods. The specific activity of the wild type strain (DK1622; O—O), the original Q4403 fusion strain (DK4368; til—CI), and the average specific activity of three independently-isolated transductants containing a single copy of pMESOO4 integrated at the Mx8 attB site (u—u) . Error bars show one standard deviation. 35 fused to lacZ (Figure 6) and integrated into wild-type M. xanthus at the attB site. Strains containing 2, 4, or 8.5 kbp of DNA immediately upstream of the (24403 fusion or 4.5 kbp of DNA normally separated by 4 kbp of sequence from the insertion were examined for developmentally regulated [3— galactosidase expression; the results are presented in Figure 7. Only 2 kbp of DNA adjacent to 94403 is required to direct developmental B-galactosidase expression. This 2 kbp segment appears to be responsible for 60% of the B— galactosidase activity driven by the 8.5 kbp region; fusion of an additional 2 kbp of upstream DNA to lacZ increases the level of expression to 75% the level observed with the 8.5 kbp segment. Transcription from promoters near the attB site is not responsible for the B-galactosidase activity as strains containing the vector alone (pREG1666) produce only a low basal level of activity (54, also see Figure 11B). This same basal level of activity is observed in strains containing 4.5 kbp of DNA normally separated from the 94403 insertion by intervening sequences (Figure 7) demonstrating that the developmental gene expression is . dependent specifically on the 2 kbp region immediately adjacent to (24403. This data suggests that the start site for transcription lies within the same 2 kbp of DNA. Additionally, the region between 2 and 8.5 kbp upstream of the 94403 insertion point has a role in regulating the level of expression. 36 A $ E §° 5 § E pMESBZ4 8:5 5:5 4 3 2 l 8 BF 1% pMESll4 l .4 pMESllS : .1 pMESliG Figure 6. Strategy for subcloning regulatory regions adjacent to 04403. (A) Additional restriction sites were located within the 8.5 kbp XhoI-BamHI segment adjacent to 04403 by digesting pMESOO3 with XhoI (X), SalI (S), ClaI (C), SacI (Sc), PstI (P), and BamHI (B), separating the products on a 0.5% TAE agarose gel, and visualizing the fragments with ethidium bromide. The approximate distance in kbp from the BamHI site is given by the number under each restriction site. (B) A 4.5 kbp XhoI-ClaI fragment, a 2 kbp PstI-BamHI fragment and a 4.0 kbp ClaI-BamHI fragment were subcloned into pREG1666 generating pMESll4, pMESllS, and pME8116 respectively. 37 50" 40" 30- I B-galactosidase specific activity (nmol ONP/mg protem'mtn) 10 - " . "/ 50 hours of development Figure 7. Developmental B-galactosidase expression from lacz fused to portions of the 8.5 kbp regulatory region upstream of 04403. Plasmids pME8114, pMESllS, pME8116, and pMESOO4 were introduced into wildtype M. xanthus at the Mx8 phage attachment site. B-galactosidase activity was measured throughout development on TPM starvation agar as described in the Materials and Methods. The specific activity of the wild type strain (DK1622; O—O), the original Q4403 fusion strain (DK4368;D-—Cl), a transductant containing a single copy of pMESOO4 integrated at the Mx8 attB site (MESOOB; A—A) and the average specific activities of five independently-isolated transductants containing a single copy of pMESll4 (O—O), pMESllS (I-I). or pMESll6 (A—A) integrated at the Mx8 attB site. Error bars show one standard deviation. 38 Despite the fact that 8.5 kbp of DNA upstream from the 94403 fusion was sufficient to direct developmentally regulated B-galactosidase expression, the activity was only 25% of the maximum level produced in the original insertion- containing strain. A similar observation was noted by Li and Shimkets (41). In their experiments, 1.3 kbp of DNA adjacent to the C-dependent Q4435 insertion was capable of directing B-galactosidase expression, but only to 50% of the maximum level seen in the original Q4435 strain. One explanation for these observations is that additional sequences further upstream are required for proper levels of expression. However, in the case of 94435, inclusion of an additional 10 kbp of upstream DNA had no effect on the expression levels. A second hypothesis is that the presence of two copies of an important regulatory sequence (one at the attB site and one at the native promoter) causes competition for a necessary factor present in limiting amounts. To test this "titration model", a control plasmid was constructed by placing the 8.5 kbp of 94403 upstream DNA into pREGll75 (25), a plasmid similar to pREG1666 but lacking the Mx8 attP segment (pME8108). This plasmid can only integrate into the Myxococcus chromosome by a homologous recombination event (Figure 3A) and places the lacz marker under the control of the native regulatory regions. If the entire promoter is contained within the 8.5 kbp fragment, expression in homologous integrants would also be expected to be reduced relative to the original insertion 39 strain since two copies of the regulatory region are present and compete for the postulated limiting factor. Figure 8 demonstrates that homologous integrants containing one copy of pMESlO8 exhibit B-galactosidase activity that is about 70% of the level seen in DK4368 (the original fusion strain). Therefore, although an additional copy of the regulatory region appears to have a small effect on gene expression, the presence of two copies of the 8.5 kbp Q4403 upstream region may not be solely responsible for the markedly reduced level of expression originally observed (Figure 6). 44 i l e a rxim l O own ram from h spar; of phe transprippipp anip ippg which i; has inserpsd, Low resolution, S1-nuclease mapping experiments were utilized to define the 5' end of the transcription unit to which (24403 was fused. A plasmid containing the 2 kbp region shown to be sufficient for directing developmental B— galactosidase expression (pMESllZ) was digested with BamHI or SalI and labelled at the 5' ends; a portion of the BamHI end-labelled probe was subsequently digested with SalI (Figure 9A and 9B, lanes 1-3. The BamHI site is located approximately 50 bp into the 5' end of Tn51ac while the SalI site lies 1 kbp upstream of the BamHI site, Figure 6). These three probes were hybridized separately to developmental RNA from Myxococcus (24 hours of development, designated T24). The hybrids were treated with S1-nuclease, 40 mm 100- B-galactosidase specific activity (nmol ONP/mg protein'min) I l ' I ' 1 0 10 20 3O 40 50 hours of development Figure 8. Developmental B-galactosidase expression in homologous integrants. A promoterless lacZ gene was fused to 8.5 kbp of DNA located immediately upstream from the 04403 insertion (pMESlO8) and introduced into wild type M. xanthus at the native chromosmal site. B-galactosidase activity was measured in cells harvested at various times during their development on TPM starvation agar as described in the Materials and Methods. The specific activity of the wild type strain (DK1622; O—O), the original Q4403 fusion strain (DK4368; CJ—D) , and the average specific activity of five independently-isolated transductants containing a single copy of . pMESlO8 integrated at the native promoter site ( I—I ). Error bars show one standard deviation. 41 Figure 9. Low resolution, 81-nuclease mapping of developmental transcripts from DK4368. Developmental (T24) or vegetative (T0) RNA (50 pg) was hybridized for 16 hours at 53° C to 5' end-labelled probe (0.5 pg) as described in the Materials and Methods. Each reaction was digested with 25 u (B) or 250 u (C) S1-nuclease. The protected products were separated on a 5% polyacrylamide-urea gel and visualized by autoradiography. Molecular size markers are indicated by horizontal bars in the margin. (A) Plasmid pMESllZ 5' end-labelled at BamHI, Sell, or labelled at BamHI and recut with SalI was used as probe. (B) BamHI*-probe (lane 1), SalI*-probe (lane 2), BamHI*-SaII-probe (lane 3), and Sl-nuclease protected products of developmental RNA hybridized to BamHI*-probe (lane 4), SalI*-probe (lane 5), or BamHI*-SaII-probe (lane 6). (C) Sl-nuclease protected products of BamHI*-probe hybridized to yeast tRNA (50 pg, lane 1), vegetative RNA (lane 2) or developmental RNA (lane 3). 42 PStI Sol] C PU ‘9 BOTH‘ P s B BomHI‘ * : : 'I probe 8 P S sott“ ‘ : : 1 probe P 8,—48 BomHI" I—-* Soil probe 123456 12 3 g es 3 43 and the protected. products were separated by gel electrophoresis. As shown in Figure 9, the same protected- fragment (about 430 bases) is observed for both the BamHI*- probe and BamHI*-SaII-probe (panel B, lanes 4 and 6); none of the SalI*-probe is protected by developmental RNA (panel B, lane 5). Furthermore, the 430 base, protected-fragment is not observed in experiments in which vegetative RNA (T0) was hybridized to the BamHI*-probe (Figure 9C, lane 2) . These observations are consistent with the idea that the 5' end of the Q4403-associated transcript is located approximately 380 bases upstream of the insertion point and also suggests that the transcript levels are developmentally regulated. ngulation pf Q44QB-assppiapsd gene expression occurs at phe level of mRNA acpumalapipn. Initial experiments indicated that the transcript fused to Q4403 was produced 24 hours after starvation but not during vegetative growth (Figure 9C). To determine whether mRNA levels were responsible for the developmental pattern of B-galactosidase activity, a quantitative Sl-nuclease experiment was performed (53). Increasing amounts of developmental RNA (T13) were hybridized to equal amounts of probe (pMESl12 digested with BamHI and labelled at the 5' end, Figure 9A); the amount of probe used was in excess (data not shown). After treatment with Sl-nuclease, the protected fragments were separated by gel electrophoresis, 44 visualized by autoradiography, and quantified using a Kodak Bio-Image densitometer. Figure 10A demonstrates that the amount of protected probe is a linear function of the amount of input RNA (at least to 100 pg of T18 RNA) and indicates the reproducibility of individual Sl-nuclease protection experiments (correlation coefficient=0.975). Identical amounts of RNA prepared from cells harvested throughout development were hybridized to a constant amount of probe and subjected to the same Sl-nuclease treatment described above. The amount of transcript present at various times during development (relative to T24 RNA) is shown in Figure 9B. For each time point, the amount of protected probe is within the linear range established in Figure 10A. The developmental increase in B—galactosidase activity correlates well with the increase in transcript abundance supporting the idea that regulation of the Q4403 fusion occurs at the level of mRNA accumulation. Whether this accumulation involves differences in transcription initiation or mRNA stability is not known. - ndn a iva ion f ne xr ion c rs 1 st parpially 1;thth DNA within 2 kbp pf tha Q4493 fpsipp, The Q4403 fusion was one of 16 Tn51ac insertions whose B-galactosidase expression absolutely depended on a functional C-signaling system (14) . This dependence is illustrated in Figure 11A. B-galactosidase expression in a csgA mutant strain containing Q4403 (DK5270) is reduced to 45 Figure 10. Quantitative 81-nuclease protection of RNA from DK4368 throughout development. (A) Upper panel: Increasing amounts of developmental RNA (T18) were hybridized to probe (0.5 pg BamHI-probe) at 53° C for 16 hours. Carrier tRNA was included with some samples to bring the total amount of RNA to 50 pg. Each sample was digested with Sl-nuclease (250 units); the protected fragments were separated on a 5% polyacrylamide-urea gel and were visualized by autoradiography. Size markers are indicated in the margin. (Lanes 1—5) 0, 12.5, 25, 50, and 100 pg T18 RNA. Lower panel: The integrated intensity of each protected band (obtained by two-dimensional densitometry on a Kodak Bio- Image densitometer) was plotted against the total amount of input RNA (correlation coefficient=0.975). (B) Upper panel: RNA (50 pg) prepared from cells harvested throughout development was hybridized to BamHI*-probe (0.5 pg) at 53° C for 16 hours. Each sample was digested with Sl-nuclease (250 units); the protected products were run on a 5% polyacrylamide-urea gel and visualized by autoradiography. (Lanes 1-6) T0, T6, T12, T13, T24, T35. Lower panel: The amount of protected product was quantitated using a Kodak Bio-Image densitometer. The integrated intensity at each time point was compared to the curve in panel A (lower) to determine whether it was in the linear range of the assay. The amount of protected probe (relative to T24;()-C» and B- galactosidase specific activity (0—0) for each time point were plotted against hours of development. 23456 i 2345 l U r 46 (ugtu,ugarord Bur/(mo Iowu) finance oggoeds esepgsotoelefi-g 200 - 100 slanaI VNUUJ enneIeJ 100 120 80 Argsuetu! pare: 5610! hours of development input RNA (pg) 47 Figure 11. Developmental B-galactosidase expression from lacz fused to the Q4403 regulatory region in csgA mutant strains. Plasmids pREG1666, pMESllS, and pMESOO4 were introduced into the csgA mutant strain DK5208 and B- galactosidase activity was measured in cells harvested throughout their development on TPM starvation agar as described in the Materials and Methods. (A) The specific activity of the wild type strain (DK1622; O—O), the original Q4403 fusion strain (DK4368; [Ii—Cl), the csgA mutant strain (DK5208; O—O), and the csgA Q4403 fusion strain (DK5270; I—I). (B) The specific activity of individual transductants containing a single copy of pREGl666 (JW103; A—A), pMEsns (11133053; O—Q), or pMESOO4 (MESOO8; Ci—Ci) integrated at the attB site. (C) The average specific activities of independently-isolated transductants containing a single copy of pREG1666 (five transductants, A—A) , pMESllS (five transductants, O—Q) , or pMESOO4 (two transductants, I—l) integrated at the Mx8 attB site in a csgA background. Error bars show one standard ‘deviation. 48 200 100- 0 4O 4 a 4 - ‘ m m 10-1 2.6.5205 95sz .065 33:8 3.8% omwemosgamé A B hours of development 49 10% of the level seen in the original insertion strain (DK4368) . In order to investigate whether the C-dependent activation of gene expression occurs through the 8.5 kbp Q4403 upstream region, lacZ fused to 0, 2, or 8.5 kbp of DNA adjacent to Q4403 was integrated into a csgA strain (DK5208) at the Mx8 attB site and developmental B- galactosidase activity was measured. The 2 kbp Q4403 upstream region was previously demonstrated to be sufficient for directing developmental gene expression (Figure 7 and Figure 11B); additional DNA within 8.5 kbp of the Q4403 fusion appeared to be involved in increasing the level of expression. Figure 11C shows that developmental l3- galactosidase activity in csgA strains containing either 2 kbp or 8.5 kbp of Q4403 regulatory DNA fused to lacz is reduced to the basal level observed in a csgA control strain containing the vector alone (pREG1666) . These results suggest that the mechanism of C-dependent gene activation is at least partially mediated through DNA within 2 kbp of the Q4403 fusion point. The data further suggests that the region between 2 and 8.5 kbp upstream of the insertion may a lso participate in C-dependent gene act ivat ion . (Alternatively, the mechanism by which the 2 to 8.5 kbp region increases gene expression may actually be C- independent with an initial requirement for C-dependent activation through only the 2 kbp of DNA immediately adjacent to Q4403). The basal B-galactosidase levels in csgA or wild-type strains containing the vector alone were 5O expected to be equivalent; however, the specific activity observed in the csgA strain is slightly higher than that of the wild-type strain. Among other explanations, overall protein or mRNA stability may be greater in csgA strains. Because the control strains show this discrepancy, the direct comparison of a particular regulatory region-lacz fusion in the csgA.:mutant. and.'wild-type strain is not possible. DISCUSSION Four signaling systems are absolutely required for M. xanthus differentiation and gene expression (3, 9, 14). Knowledge about the C-signal system, particularly the signal molecule itself (C-factor) is rapidly increasing (10, 28, 29, 30). To begin to define cis-acting elements involved in C-dependent gene activation DNA adjacent to developmentally- regulated Tn51ac fusions was cloned and tested for promoter activity. Although DNA within 1.8 kbp of the Q4414 insertion is unable to direct developmental B-galactosidase expression (Figure 4) , a 2 kbp region immediately adjacent to the Q4403 insertion is sufficient to promote B- galactosidase activity (Figure 7). Interestingly, the level of activity increases when an additional 2 or 6.5 kbp of Q4403 upstream DNA is fused to the lacZ marker gene. This observation suggests that several different segments of upstream DNA are involved in activating Q4403-associated gene expression. Similar activating sequences have been identified in the developmentally regulated genes mbhA (55), ops, and tps (57, 58). A cis-acting region of DNA located between 89 and 276 nucleotides upstream of the mbhA transcription start site is required for the accumulation of mbhA transcripts during development and DNA further than 2 kbp from the transcription start site is also believed to 51 52 play a role in gene expression (55). Expression from lacZ fused to DNA upstream of the Lops gene increases substantially when a region between -131 to -208 bp is included (56). This same ops segment functions as an upstream activating site (UAS) for transcription of the tps gene (located about 2 kbp downstream of the UAS, 57) Together, this evidence indicates that DNA separate from the transcription start site plays an essential role in regulating M. xanthus developmental gene expression. The specific mechanisms coupling intercellular signaling to gene activation during M. xanthus development are currently unknown; however, by comparison to other prokaryotic and eukaryotic organisms which undergo similar developmental cycles, regulation of transcription initiation is likely to be involved. The start site of the Q4403- associated transcript was located about 380 bases upstream of the insertion point (Figure 9). Using quantitative Sl- nuclease experiments, the levels of transcript were demonstrated to be developmentally regulated (Figure 108) . These results provide preliminary evidence that modulation of gene expression occurs at least partly through the regulation of transcription initiation or mRNA stability. In the case of the Q4403-associated transcription unit, the mechanism of gene activation is clearly dependent on a functional C-signaling system as B-galactosidase expression from the original Tn51ac insertion is essentially abolished in a csgA mutant (Figure 11A; [14]) . This C-dependent 53 activation is mediated to at least some extent through the 2 kbp Q4403 upstream region (Figure 118 and 11C). Perhaps in response to C-signaling, trans-acting factors are produced or already existing factors are modified which interact with this region to stimulate transcription. The exchange of C signal could also trigger the degradation of a repressor normally blocking the transcription of the Q4403-associated gene. Expression of another C-dependent gene, ops, requires the product of the sigB locus (a putative developmental a factor, 58) and a DNA binding activity that preferentially associates with the ops UAS was identified in developmental M. xanthus crude extracts (57). It will be very interesting to biochemically identify the trans-acting factors that are involved in C-dependent gene activation. Further characterization of the Q4403-associated regulatory regions including a more precise mapping of the start site, sequence analysis of the promoter, and localization of the upstream activation sites within the 8.5 kbp adjacent to Q4403 will facilitate the identification of individual components of the transcriptional machinery. The maximum level of B-galactosidase expression observed in transductants containing up to 8.5 kbp of Q4403 upstream DNA fused to the lacz marker was only about 20% to 30% of the level produced in the original fusion strain (Figures 5 and 7). The lower activity level was not due to the titration of a limiting factor through competition between binding sites at the native and putative promoters 54 since strains containing' both .lacz fused to the native promoter as well as a copy of the putative promoter approximately 20 kbp downstream produce B-galactosidase levels comparable to the original Q4403 fusion strain (Figure 8). A possible explanation for the decreased [3- galactosidase expression is that additional regulatory sequences located further than 8.5 kbp from the fusion site may be required for maximal gene expression. Since the transcription start site is located about 380 bases from the 5' end of the insertion, these putative, additional regulatory regions would have to be strong activating sequences (as they would be responsible for 60% to 70% of the B-galactosidase activity observed in the original Q4403 fusion strain) and would have to function over a substantial distance (greater than 8 kbp). Although eukaryotic enhancer elements are often located within a few hundred base pairs of the transcription start site, long-range activation sequences which function at distances greater than several kilobases upstream or downstream of the transcription start site have been identified in several systems (59, 60, 61). The M; xanthus ops UAS also affects the expression of the tps gene located 2 kbp downstream (57) suggesting that long- range activating sequences might be involved in Myxococcus developmental gene expression. The possibility also exists that regions downstream of the start site may affect gene activation. C-dependent gene expression may be novel in its requirements for DNA extremely distant from the 55 transcription start site. Confirmation of these hypotheses requires cloning additional Q4403 upstream or downstream regions and testing them for transcription-enhancing activity. Several alternative explanations for the lower B- galactosidase activities can be postulated including a positional effect on gene expression as well as autorepression by the Q4403-associated gene product on its own transcription. Besides being fused to the entire complement of native promoter sequences, the original Tn51ac insertion is at a completely different chromosomal location with respect to the Mx8 attB site. Variations in the levels of gene expression which are solely dependent on chromosomal position have been documented in both eukaryotic and prokaryotic systems including M. xanthus (56, 62, 63). 131 the current studies, the attB site may be a particularly poor location for C-dependent gene expression. The position- effect model can be directly tested by integrating the regulatory region-lacz fusions at a different chromosomal location, such as near the original locus. Position effects would be eliminated and B-galactosidase levels should approach those observed in the original fusion strain. The reduced levels of B-galactosidase expression may, however, reflect the true regulation of the Q4403- associated transcript. If the original Q4403 Tn51ac inserted within a gene whose product negatively affects its own express ion , B-galactosidase activity would be 56 artificially elevated in the insertion strain due to the absence of the autorepressor product. Transductants containing the Q4403 upstream region fused to lacz and integrated at the attB site would still produce the repressor molecule from the native locus and thus be subjected to autoinhibition. The autorepression model can be investigated by introducing the regulatory region-lacZ fusions into a modified insertion strain possessing a Tn5- Tcr replacement of Tn51ac-Kmr. If this model is correct, the ldevelopmental B-galactosidase levels should be comparable to the original Tn51ac Q4403-containing strain. Alternatively, a comparison of the levels of the Q4403-— associated transcript in the wild-type strain (DK1622) and the Tn51ac fusion strain (DK4368) could be performed; the autorepressor model predicts that the transcript levels would be 3- to 5-fold higher in DK4368 than in DK1622. The results of these proposed experiments should yield valuable information that will allow a more definitive explanation for the decreased B-galactosidase expression. At the same time, this information 'should provide insight into increasing the window of sensitivity for detecting changes in the level of B-galactosidase activity which is necessary for a more detailed analysis of the cloned regulatory regions. CHAPTER 2 CHARACTERIZATION OF VEGETATIVE RNA POLYMERASE FROM MYXOCOCCUS XANTHUS One mechanism utilized by a variety of bacteria to coordinate proper sequential gene activation with environmental conditions is the modification of the promoter specificity of RNA polymerase (32). Promoter recognition is mediated by a sigma (0) subunit which binds to both the RNA polymerase core (a2, [3, B') and to the template's promoter. Nitrogen regulation and the heat shock response in E. coli are regulated by sigma factors (0'54 and 0'32 respectively; 64, 65) with promoter specificities quite different from the major E. coli 6 factor (670). In B. subtilis, the endospore developmental program induced by nutrient depletion is tightly coupled to the production or activation of alternative 0' factors (32) . Similarly, heterogeneous RNA polymerase molecules discovered in Streptomyces coelicolor may be involved in differential gene expression during mycelial and hyphal development (66, 67). During differentiation, M. xanthus exhibits dramatic changes in. both. protein. synthesis and. patterns of gene expression (12, 68) suggesting the need for mechanisms to effect these tightly regulated changes. A comparison of four well-characterized promoters from .M. xanthus genes which are utilized during growth (vegA, [69]) or development 57 58 (ops and tps [70, 71], mbhA [55]) reveals differences indicative of polymerase heterogeneity. Three of the promoters (vegA, ops, and tps) share weak homology with the -10 and/or -35 sequences of the E. coli 0'70 consensus sequence. The promoter of the remaining gene (mbhA) is clearly distinct as it resembles an E. coli 0'54 consensus sequence (72). The variations indicate that modification of the promoter-recognition specificity of RNA polymerase must occur to allow utilization of these distinct promoters. Indeed, the gene for a development-specific 0' factor was recently cloned (sigB, 58). This new 0' factor is expressed during middle to late development and is required for proper maturation of myxospores. Deletions of the sigB gene do not affect the production of protein S (tps gene product). However, the sigB deletions abolish the production of protein S1 (ops gene product), suggesting that the sigB-encoded factor directly interacts with the ops promoter or indirectly affects its expression by interrupting the program of developmental gene expression. Evidence, therefore, is accumulating to support the theory that differentiation in M. xanthus is coordinated by the regulated expression and utilization of alternative 0' factors. Intercellular signaling is absolutely required for developmental gene expression in M. xanthus and modulation of transcription initiation may be at least one mechanism coupling the two. An in vitro system for reconstituting 59 signal-dependent transcription will ultimately be required to test this hypothesis. Chapter 1 of this study described initial characterization of C signal-dependent promoter and regulatory regions in M. xanthus; these promoters could serve as templates for in vitro reconstitution experiments. In addition to a repertoire of well characterized C- dependent templates, several proteins are required. Included among these proteins are core RNA polymerase, developmental 6 factors, and additional regulatory proteins that might act at upstream (or downstream) sites and be necessary for efficient transcription. In preparation for the eventual establishment of an M. xanthus in vitro transcription system, core RNA polymerase from M. xanthus was purified for reconstitution with developmental factors. The availability of vegetative promoters from several bacteria prompted initial purification attempts with vegetative M. xanthus as a model system. Using a modified Burgess-Jendrisak procedure (73, 74), both core and holoenzyme were obtained. A protein was isolated from holoenzyme fractions that stimulated both M} xanthus and En coli core polymerase transcribing activity. Additionally, the transcriptional activity of holoenzyme on a variety of vegetative and developmental templates was examined. 60 MATERIALS AND METHODS Grpwph pf strains Vegetative Myxococcos xanthus DK1622 cells were grown in six liters of CYE (1% Casitone [Difco Laboratories], 0.5% yeast extract, 0.1% MgSO4) with shaking (250 rpm) at 30°C to the mid-exponential phase of growth (7.5 x 108 cells/ml). Cells were harvested by centrifugation in a GS-3 rotor (7500 rpm, 10 minutes), quickly frozen in a dry ice-ethanol bath, and stored at -70°C. Approximately 80 grams of cell (wet weight) were obtained. Baffers Buffers and other solutions were prepared as described previously (73, 74). Grinding buffer contained 0.05 M Tris- HCl (pH 7.9), 10% (v/v) glycerol, 1 mM EDTA, 0.2 mM DTT, 1 mM 2-mercaptoethanol, 130 pg/ml lysozyme, 0.23 M NaCl, 1 mM PMSF, and Trasylol (106 KIU/L, Boehringer Mannheim Biochemicals). The buffer ‘used. throughout. most of the isolation was TGED (0.1 M Tris-HCl [pH 7.9], 10% glycerol, 1 mM EDTA, 0.2 mM DTT) or TGEDM (TGED buffer with 10 mM MgC12) containing sodium chloride as mentioned. Ammonium sulfate dilution buffer consisted of 40 mM Tris-HCl (pH 7.9), 1 mM EDTA, 0.2 mM DTT, and 65% saturated ammonium sulfate. Samples were dialyzed into storage buffer (0.1 M Tris-HCl [pH 7.9], 50% glycerol, 0.1 M NaCl, 1 mM EDTA, 0.3 61 mM DTT, and 10 mM MgC12) and kept at -20° C as indicated. RNA polymerase purificapion RNA polymerase was purified from MYxococcus using procedures described by Jendrisak and Burgess (74) and Rudd and Zusman (73). The entire purification was performed rapidly and in the cold (10°C). More specifically, 80 grams of vegetative cells were resuspended in 240 ml of grinding buffer by blending at low speed for 2 minutes in a Waring blender. After letting the solution sit for 20 minutes in an ice-water slurry, sodium deoxycholate (4%) was added to a final concentration of 0.05%. The mixture was blended for 30 seconds at low speed and again set in the ice-water slurry for 20 minutes. The viscous solution was sheared at high speed for 60 seconds, diluted with 320 ml TGEDM buffer containing 0.2 M NaCl, and blended for 30 additional seconds at low speed. Cell debris was pelleted by centrifugation in a GSA rotor (8000 rpm, 45 minutes). The supernatant was collected and a portion retained as the "low speed supernatant". Protein was precipitated by slowly adding 10% Polymin P (polyethyleneimine [Sigma Chemical Co.] prepared as detailed in [75]) to a final concentration of 0.25%. The slurry was stirred for 5 minutes, spun in a GSA rotor (6000 rpm, 15 minutes), and the liquid phase removed. The Polymin P pellet was resuspended in 320 ml TGEDM buffer containing 0.35 M NaCl (using a homogenizer) and was washed for 5 62 minutes with gentle stirring. After centrifugation (6000 rpm, 15 minutes), the supernatant was discarded and RNA polymerase was eluted from the Polymin P by resuspending and stirring the pellet. in. TGEDMI containing 1.0 M’ NaCl as outlined above. The Polymin P was removed from the suspension by spinning (8000 rpm, 30 minutes), and the yellow supernatant was saved. A fractional ammonium sulfate precipitation was performed to remove additional contaminating proteins. Powdered ammonium sulfate was slowly added to 40% saturation (over a period of 20 minutes) with stirring for another 20 minutes. The precipitated proteins were collected by centrifugation in a GSA rotor (8000 rpm, 45 minutes) and discarded. RNA polymerase was recovered from the liquid phase by increasing the ammonium sulfate to 65% saturation, stirring for' 20 :minutes, diluting the ‘mixture with 65% saturated-ammonium sulfate dilution buffer, and spinning at 8000 rpm for 45 minutes. The dark, yellow pellet was dissolved in 87 ml TGED buffer and an aliquot saved as the "ammonium sulfate enzyme". The ammonium sulfate enzyme was applied to a 30-ml double-stranded DNA-cellulose affinity column (7mg/ml calf- thymus DNA [Sigma Chemical Company], equilibrated with TGED containing 0.15 M NaCl) at a rate of 43 ml/hour. RNA polymerase was eluted with a 0.15 M to 1.3 M NaCl gradient in 172 ml TGED buffer at the same flow rate. Two-ml fractions were collected and assayed for transcribing 63 activity. Fractions with polymerase activity were dialyzed against 3 liters of storage buffer for 4 hours (one change after 1 hour) and stored at -20°C. In vipro transcriptipn assays Polymerase activity was initially located using a nonspecific transcription reaction. The standard assay conditions were similar to those used in the procedure of Rudd and Zusman (73) except 4 pg poly d(AT) was utilized as template, 0.1 mM each ATP and UTP were included as the unlabeled ribonucleotides, 2 pCi 3H-UTP (New England Nuclear) was the labeled ribonucleotide, and KCl was omitted. Ten microliters of the samples to be tested were added to each reaction and the subsequent incubation, precipitation, and. wash steps were followed exactly as described (73). RNA polymerase-containing fractions were analyzed for their ability to produce run-off transcripts on linearized plasmids possessing one of several bacterial promoters. Reaction conditions were those of Kroos et. a1. (76). For reconstitution experiments, 2 pl of core RNA polymerase were mixed with 8 pl of gel-purified protein (77) and incubated on ice for 10 minutes. Ten microliters of the fractions to be tested, or reconstituted enzyme, were included in each assay. E. coli c70-RNA polymerase was a gift from A. Revzin; E. coli core RNA polymerase was a gift from C. Gross; B. subtilis 643-RNA polymerase was purified by L. Kroos. 64 SDS-pplyappylamide gals Core polymerase subunits (a, B, B') and total protein were separated using SDS-polyacrylamide gel electrophoresis. Aliquots from fractions were mixed with one-third volume sample buffer (0.375 M Tris-HCl [pH 6.8], 6% SDS, 15% 2- mercaptoethanol, 30% glycerol, 0.3% bromophenol blue) and boiled 2 minutes. The samples were loaded onto a 10% discontinuous SDS-polyacrylamide gel and were separated by electrophoresis at a constant voltage (100-200 V) for several hours. Proteins were visualized by staining in a Coomassie blue 2R250 solution. (50% .Methanol, 7.5% acetic acid, 0.1% Coomassie blue R250) for 15-30 minutes, followed by rapid destaining (10% EtOH, 7.5% acetic acid) with four 15 minute rinses. Alternatively, a more sensitive staining was performed using a Bio-Rad silver stain kit according to the manufacturer's specifications. When proteins were to be gel-purified, total protein was precipitated with one volume of 20% TCA, washed with one volume 5% TCA, and dried 5 minutes at room temperature. The pellets were resuspended. in 1x sample buffer, boiled 2 minutes, and loaded onto a preparative gel. The desired bands were sliced. out. of ‘the Coomassie-stained gel and recovered by the procedure of Hager and Burgess (78) for use in reconstitution experiments as mentioned. W t rn lo an 1 is Proteins separated on SDS-polyacrylamide gels (as described above) were transferred to PVDF membranes using 65 the procedure of Matsudaira (78). The blots were incubated in TBS blocking buffer (20 mM Tris-HCl, [pH 7.9], 500 mM NaCl, and 2% nonfat dry milk) for 2 hours at room temperature to prevent nonspecific binding of the antibodies. The membranes were shaken overnight at room temperature in a 1:600 dilution of polyclonal antiserum raised against B. subtilis 0'43 (gift of R. Doi) in antibody buffer (TBS containing 2% nonfat dry milk and 0.05% TMeen 20). Immunodetection using a secondary goat anti-rabbit antibody conjugated to alkaline phosphatase was performed according to the manufacturer's specifications (Bio—Rad). RESULTS Earificapion pf vegetative RNA polymerase by DNA—cellulose phrpmatography. Using a modified version of a polymerase purification scheme described by Jendrisak and Burgess (73) and Rudd and Zusman (74), RNA polymerase was purified from vegetative M. xanthus. After lysing the cells and pelleting debris by low speed centrifugation, a supernatant is obtained that contains non-specific transcribing activity on a poly d(AT) template (300-850 cpm under the conditions in Figure 2). This low speed supernatant is incapable of producing run-off transcripts from the B. subtilis veg promoter which has the E. coli 0'70 consensus at its -35 and -10 regions (79). Furthermore, polymerase subunits (or, B, B', 0') are also not easily distinguishable from the many other contaminating proteins on Coomassie blue-stained SDS-polyacrylamide gels (Figure 1, LS). Following additional purification steps involving Polymin P precipitation and ammonium-sulfate fractionation, core polymerase subunits are visible on Coomassie blue-stained SDS-polyacrylamide gels (Figure 1, AS) and non-specific transcribing activity increases 2- to 3-fold (900-1200 cpm under the conditions in Figure 2). Additionally, the ammonium-sulfate enzyme is capable of utilizing the Bacillus veg promoter to produce run-off 66 67 Figure 1. SDS-polyacrylamide gel analysis of proteins in various fractions of the purification. Proteins in the low speed supernatant (5 pg), ammonium sulfate enzyme (3.7 pg), and DNA-cellulose fractions (25 pl) adjacent to or containing the poly d(AT) peak (Figure 2) were separated on 10% SDS-polyacrylamide gels and visualized by Coomassie blue staining. The numbers above each lane indicate the fraction; low speed supernatant (LS), ammonium sulfate enzyme (AS). E. coli polymerase subunits (01, B, B', o) are indicated as size markers in the left margin. 68 69 transcripts of the appropriate size. However, the background of both shorter and longer transcription products is quite high (data not shown). The ammonium-sulfate enzyme was chromatographed on a double-stranded DNA-cellulose column as described in the Materials and Methods. A narrow peak (Figure 2, fractions 45-51) followed by a small tail (Figure 2, fractions 52-56) of poly d(AT) activity elutes within a 0.45 M to 0.52 M NaCl gradient. Fractions across the gradient, particularly those near or containing the peak of poly d(AT) activity, were examined for their protein composition by SDS- polyacrylamide gel electrophoresis (Figure 1) and were also tested for the production of run-off transcripts on the B. subtilis veg promoter (Figure 3). Core polymerase is detectable in fractions 40-60 with the highest levels appearing in fractions 48-51 (Figure 1). These latter fractions are also coincident with the peak of non-specific transcribing activity (Figure 2). In contrast, fractions containing the "tail" of poly d(AT) activity (Figure 2, fractions 52-56) have 2- to 10-fold less core polymerase subunits visible on SDS-polyacrylamide gels. Interestingly, the levels of core polymerase subunits do not always correlate with the ability to produce specific transcripts from the veg promoter. Fractions eluting early in the poly d(AT) peak (Figure 2, fractions 46 and 47) that have amounts of a,|i and B' comparable to those in the "tail" fractions (Fractions 54 and 52, respectively) produce at least 10-fold 70 2000 0.8 i- ,l ;’ ~o7 E o. O. .0 e .0 .>.~ " 0-6 E ,o‘ ,4 A .‘ " 0 5 c t— . , ,g s. _ i i S 'O '0 .. 2 1 000 "‘ I '- 0.4 8 ° 2 °‘ 8 o 3 0.3 F3, 3 2 >4 ‘6 - 0.2 o < l E r- 0.1 0 . 1 . u . r . 0 0 0 2 0 4 0 6 0 8 0 fraction number Figure 2. DNA-cellulose chromatography. The ammonium sulfate enzyme was dissolved in TGED buffer (87 ml) and applied to a 30-ml double-stranded, calf-thymus DNA-cellulose column as described in the Materials and Methods. RNA polymerase was eluted with a 0.15 M to 1.3 M linear NaCl gradient (O--O) . Fractions (2 ml) were collected. and assayed for incorporation of 3H-UTP into RNA with poly d(AT) as template (O--O) . (A void volume of one-third the column volume was assumed). 71 Figure 3. Run-off transcription by vegetative RNA polymerase from M. xanthus and B. subtilis. (A) DNA- cellulose fractions (10 pl) across the poly d(AT) activity peak (Figure 2) were tested for the production of run-off transcripts as described in the Materials and Methods. BamHI-digested pMSS30 containing the B. subtilis veg promoter was utilized as template. Only the region of the gel containing the run-off transcripts is shown. Numbers above each lane indicate the fraction being tested. (B) Run- off transcripts from the B. subtilis veg promoter were produced by B. subtilis (lanes 1-3) and M. xanthus (lanes 4-6; fraction 54 from Figures 1 and 2) DNA-cellulose- purified RNA polymerase. Template (pMSSBO) was digested with either BamHI (lanes 1 and 4), EcoRV (lanes 2 and 5), or HindIII (lanes 3 and 6). Size markers are indicated in the margin. 72 A 42444546474849505152535455565860 —ll0 O t _. 73 lower levels of specific transcripts. These results suggest that fractions eluting early in the salt gradient, which. contain. both fairly’ high levels of a, B, and B' subunits and poly d(AT) activity but low specific transcribing activity (Fractions 46-48), are o-depleted, core RNA polymerase. Fractions eluting later in the gradient which have higher levels of specific transcribing activity despite lower amounts of a" B, and B' and ’poly d(AT) activity (Fractions 52-56), are o-enriched, holoenzyme- containing samples. To ensure ‘that ‘transcription.‘was initiating at the promoter and proceeding to the correct end of the template, a plasmid containing the Bacillus veg promoter was digested with several restriction. enzymes and. was utilized' as a template in transcription reactions. DNA-cellulose-purified B. subtilis and M. xanthus RNA polymerase (fraction 54) produce identical transcripts of the expected sizes on the Bacillus veg promoter (Figure 3B) indicating that transcription is in the correct direction. (Two bands are sometimes seen when BamHI-digested veg template is used in transcription reactions with either Bacillus or Myxococcus RNA polymerase; the identity of the second band is unknown. The lower band could be a degradation product of the larger transcript or different termination or initiation sites could be used to generate the two products.) 74 M, yanphus core polymerase astivity can be stimulated by E. £911 g 1Q; Several fractions from the DNA-cellulose column (fractions 46-48; Figures. 1, 2, land 3) *which showed a reasonable abundance of polymerase subunits 01, B, and B' were capable of nonspecific transcription. Because these fractions produced fewer run-off transcripts from the Bacillus veg promoter, they were postulated to contain core RNA. polymerasel To test this hypothesis, E. coli 670 purified from an SDS-polyacrylamide gel and renatured by the Hager-Burgess procedure (77) was used with the putative M} xanthus core RNA polymerase in transcription reconstitution experiments. Figure 4 clearly shows that the transcriptional activity of putative M. xanthus core polymerase on the Bacillus veg promoter (Fraction 47, lane 1) is increased 2- to 3-fold by 07° from E. coli (lane 2). The amount of stimulation is identical to that observed for E. coli core polymerase (lane 4) upon supplementation with gel-purified 670 (lane 5). Additional M. xanthus fractions (46 and 48) are similarly stimulated [by E. coli 0'70 (data not shown). These observations indicate that fractions 46-48 obtained by DNA-cellulose chromatography contain o-depleted, M. xanthus core RNA polymerase which can function with a heterologous sigma factor to produce specific transcripts. 75 Figure 4. stimulation of E. coli and M; xanthus core RNA polymerase transcriptional activity. Run-off transcripts from the B. subtilis veg promoter were produced using reconstituted E. coli and M. xanthus polymerase. Core RNA polymerase (2 pl of E. coli core or M. xanthus fraction 47 from Figure 1) and either gel-purified E. coli 670 (8 p1), gel purified M. xanthus 50 kDa protein (8 pl), or guanidine dilution buffer (8 pl, as a control) were mixed and incubated on ice for 10 minutes prior to the addition of template. Transcription reactions were performed as described in the Materials and Methods. The products were separated on a 5% polyacrylamide-urea gel and visualized by autoradiography. M. xanthus core RNA polymerase (fraction 47) + control buffer (lane 1), E. coli 670 (lane 2), or M. xanthus 50 kDa protein (lane 3); E. coli core RNA polymerase (fraction 47) + control buffer (lane 4), E. coli 670 (lane 5), or M. xanthus 50 kDa protein (lane 6). Only the region of the gel containing the run-off transcripts is shown. 76 .M; xanthus holoenzyme can utilize different bacterial promotsrs. DNA-cellulose fractions that generated run-off transcripts on the Bacillus veg promoter were tested for their ability to 'utilize other bacterial promoters. .A comparison of the -35 and -10 sequences of the promoters used in this study to the E. coli 670 consensus is shown in Table 1. M. xanthus holoenzyme (fraction 52) produces run- off transcripts of the appropriate size from the E. coli lacUV5 promoter (80) as efficiently as it does from the B. subtilis veg promoter (data not shown). Both the lacUV5 and veg promoters have -35 and -10 sequences that are (within one base pair) identical to the E. coli consensus sequences. In contrast, the M. xanthus vegA promoter (69) which has very low sequence similarity to the E. coli consensus sequence, is unable to direct transcription using M. xanthus polymerase fractions (fractions 40-60 from Figure 2, data not shown). DNA-cellulose-purified B. subtilis 643-polymerase and E. coli holoenzyme are also incapable of producing transcripts from this vegA promoter. DNA-cellulose fractions across the gradient were tested for the production of transcripts from an M. xanthus, A signal-dependent, developmentally-regulated promoter. A plasmid containing 300 bp of DNA upstream of the A-dependent Q4521 insertion (which has been shown to be sufficient for directing developmental gene expression, 81) was utilized as template. No specific transcripts are produced by vegetative polymerase (fractions 42-59 from Figure 1 were Table 1. 77 Comparison of several bacterial promoters. -35 -10 E. coli 0' “’0 consensus TTGACA TATAAT B. subtilis veg TTGACA TACAAT E. coli lacUV5 TTTACA TATAAT M. xanthus vegA TAGACA AAGGGT M. xanthus tps TTGCAT AATGCT M. xanthus ops TTGCTC TCTGCT ML xanthus mbhA TTGGCA N5 TCTGCT E. coli 054 consensus CTGGCA N5 TTTGCA Table 1. Comparison of several bacterial promoters. A comparison of the -10 and -35 sequences of the bacterial promoters utilized in this study to the E. coli 670’ consensus sequence is shown. The three, well-characterized developmental promoters from M. xanthus and the E. coli 054* consensus sequence are also listed. 78 tested, data not shown). A 50 kDa protein present in M. xanthus holoenzyme containing fracpions is capabla of stimulaping transcription, Previous studies have suggested that Myxococcus vegetative RNA polymerase is associated with a 0' factor(s) having a molecular weight of 70 to 80 kDa. Rudd and Zusman (73) demonstrated that two proteins (CI and GII) present in vegetative M. xanthus migrated slightly higher than E. coli 070 on SDS-polyacrylamide gels and consistently copurified with RNA polymerase activity. Additionally, the gene for a putative vegetative 0' factor was cloned from Myxococcus based on homology to the E. coli rpoD gene (82), and sequence analysis subsequently predicted two possible protein products of 73 kDa and 80 kDa molecular weights. Analysis of the subunit composition of holoenzyme-containing fractions from this study by SDS-polyacrylamide gel electrophoresis and Coomassie blue-staining fails to reveal an obvious 0' factor in the 70 to 80 kDa range (Figure 1, fractions 50-55). A more sensitive, silver stain was used to examine core and holoenzyme fractions (Figure 5). Although several bands in the 65—95 kDa range are faintly visible in holoenzyme fractions (fractions 50-55), none of the proteins is as abundant as would have been predicted from the high levels of specific transcribing activity or the amounts of or, B, and B' present. Furthermore, all of these proteins appear to be present in core fractions as well. 79 50HD-' Figure 5. Comparison of proteins present in fractions containing M; xanthus core and holoenzyme. Proteins in DNA- cellulose fractions across the peak of specific transcribing activity (20 pl, Figure 3) were separated on a 10% polyacrylamide gel and visualized by silver staining. The numbers above each lane indicate the fraction; one microgram of E. coli RNA polymerase was used as a marker. The putative 50 kDa transcription-stimulating protein is indicated in the margin. 80 To facilitate the identification of a 6 factor, antibodies raised against Bacillus 7643 (a vegetative 6 factor that recognizes promoters with E. coli 670 consensus sequences) were used in Western blot analysis of DNA- cellulose-purified fractions. The anti-6'43 antibodies recognize many proteins in both Bacillus vegetative extracts (Figure 6B, lane 1) and in E. coli holoenzyme (Figure 6B, lane 2) and react quite strongly with Bacillus 0'43 and E. coli 670. Similarly, most of the .M. .xanthus proteins visible on a Coomassie blue-stained blot (Figure 6A, lanes 3 and 4) are also .recognized. by the polyclonal antiserum (Figure GB, lanes 3 and 4). However, two proteins present only in the holoenzyme fraction (fraction 50) which are undetectable on the Coomassie-stained blot (Figure 6A, lane 4) are recognized by the anti-o43 antibodies (Figure 6B, lane 4; indicated in the margin). These two proteins migrate with molecular weights of approximately 50 kDa and 70 kDa and are not detectable in the core fraction (fraction 47, Figure 6B, lane 3) suggesting they may be candidates for a 6 factor. A preparative SDS-polyacrylamide gel was used to separate proteins in holoenzyme-containing fractions (Figure 7A). The entire samples from fractions 52 and 53 were precipitated and loaded into one lane of the gel as described in the figure legend. Several slices in the 70-90 kDa range as well as one in the 50 kDa range were cut from the gel. The proteins within these slices were 81 M 00 b M 03 A —70| B c \/ I Figure 7. Isolation of an M. xanthus transcription stimulating protein. (A) A holoenzyme-containing fraction (1 ml of pooled DNA-cellulose fractions 52 and 53, Figure 2) was TCA-precipitated, the proteins were separated on a 14-cm 10% SDS-polyacrylamide gel and were visualized by Coomassie blue staining (lower lane). E. coli holoenzyme (1 pg) was used as a marker (upper lane). The direction of migration through the gel is indicated by an arrow. M. xanthus proteins were subsequently purified from the indicated gel slices (1 mm) and used in reconstitution experiments. (B) Run-off transcripts from the B. subtilis veg promoter were produced with reconstituted M. xanthus polymerase. M. xanthus core polymerase (2 pl of fraction 47, Figure 1) and either guanidine dilution buffer (8 pl as a control, lane C) or gel-purified protein (8 pl from the corresponding gel slices in panel [A]) were mixed and incubated on ice for 10 minutes prior to the addition of template (pM8530 digested with BamHI) . Transcription reactions were performed as described in the Materials and Methods. The products were separated on a 5% polyacrylamide-urea gel and were visualized by autoradiography. Only the region of the gel containing the run-off transcript is shown. 83 eluted, denatured, and renatured using the Hager-Burgess procedure (Figure 7A). When proteins from gel slices in the 70-90 kDa range are mixed with M. xanthus core polymerase and are tested for the production of run-off transcripts from the B. subtilis veg promoter, no stimulation of transcription is observed (Figure 7B). These results indicate that either the concentration of proteins eluted from the gel slices in the 70-90 kDa range is too low to stimulate transcription or that none of the 70-90 kDa proteins is the missing 6 factor(s). In contrast, a protein which was eluted from the 50 kDa gel slice has a strong transcription-stimulating activity (Figure 7B). This 50 kDa protein is capable of stimulating both M. xanthus and E. coli core polymerase about 20-fold (Figure 4, lanes 3 and 6) on the Bacillus veg promoter, but is unable to stimulate B. subtilis core (data not shown). Additionally, upon re-examination of a silver-stained SDS- polyacrylamide gel of proteins across the elution gradient (Figure 5), a protein clearly present in holoenzyme- containing fractions (fraction 50-55) that migrates at approximately 50 kDa is less abundant or absent in early core fractions (Figure 5, fractions 46 and 47 indicated in the margin). Whether this protein is the 50 kDa protein that stimulates transcription or merely migrates to the same position is unknown. Overall, these results suggest that a 50 kDa M. xanthus protein may be a vegetative 6 factor. DISCUSSION Regulation of transcription initiation by the use of alternative 0' factors has been postulated as a mechanism involved in developmental gene expression during M. xanthus differentiation. Initial purification of RNA polymerase from vegetative M. xanthus was attempted to obtain core RNA polymerase for use in future reconstitution experiments aimed at identifying developmental transcription factors and to gain an appreciation for the difficulties that might be encountered in the isolation of developmental RNA polymerase. RNA polymerase was purified from vegetative M. xanthus using a relatively simple procedure (73, 74) including’ double-stranded. DNA-cellulose Chromatography; Roughly 1.8 mg of total polymerase (core and holoenzyme) was obtained from 80 g of cells. This yield was lower than that reported by Rudd and Zusman (73 at the same point in their purification). Much of this difference can be attributed to the overall lower levels of protein consistently observed in the low speed supernatant fraction. The low yield may increase the difficulty of purifying developmental polymerase since large amounts of starting material are not easy to obtain. By washing the cells prior to lysis to remove extracellular proteases, achieving better cell lysis (perhaps by including a sonication step), and decreasing the 84 85 initial centrifugation speed used to prepare the low speed supernatant, yields may be increased.- Polymerase subunits 01, B, and B' are clearly the most abundant proteins by SDS-polyacrylamide gel analysis after the DNA-cellulose column. No easily identifiable o-factor is present (Figure 1) , suggesting that most of the polymerase obtained is core RNA polymerase. However, all fractions are contaminated with some 6 as at least low levels of run-off transcripts from the Bacillus veg promoter are produced even by "core" fractions (Figure 3, fractions 46-48). The contaminating proteins (including 6) can possibly be removed by chromatography on a phosphocellulose (83) or Bio-Rex 70 column (74); at low salt concentration, core RNA polymerase binds to these columns while 0 may be released into the flow-through. Alternatively, core RNA polymerase can be separated from holoenzyme by single-stranded DNA cellulose chromatography using a salt step-elution (74, 85) in combination with Bio-Rex chromatography. The transcriptional activity of Myxococcus core polymerase on the veg promoter can be stimulated by supplementation with either gel-purified 0’70 or a gel- purified 50 kDa M. xanthus protein (Figure 4) . Reconstitution of developmental transcription may not be as simple. DNA located up to 2 kbp from the transcription start site of several development-specific genes (ops and tps [56, 57], mbhA [58]) appears to be involved in the regulation of their expression. Furthermore, gel mobility 86 retardation experiments have demonstrated the presence of developmental DNA binding proteins in crude cell extracts of M. xanthus which recognize an ops upstream activation sequence (57). These observations suggest that additional developmental activators and repressors may complicate transcription reconstitution experiments. Nonetheless, vegetative core RNA polymerase provides a convenient starting point upon which to build the additional transcription machinery. The gene for the major vegetative 0' factor in M. xanthus (sigA) has been cloned based on its homology to the rpoD gene of E. coli (82). In amino acid sequence comparisons, the carboxy-terminal domain of the Myxococcus d-factor shares 78% similarity with E. coli 670. In particular, region 2 which is involved in core binding and recognition of the -10 sequence shows only 3 conservative changes between these two 0 factors. As predicted by the structural analysis, vegetative RNA polymerase purified from M. xanthus recognizes promoters with the E. coli 070 consensus sequence (Figure 5). On the other hand, neither the vegA promoter from MYxococcus (69) which conforms only weakly to the same consensus sequence (specifically at the - 1 0 region) nor the A-dependent , Q4 52 1 -associated devlopmental promoter (81) is able to serve as template. These observations can potentially be explained by several hypotheses. The amount of o-saturated enzyme obtained at this point in the purification may not have been sufficient 87 to stimulate transcription from these promoters. Alternatively, a minor vegetative 0 factor, a developmental Ofactor (in the case of the Q4521-associated promoter), or additional activators may be needed. Finally, a repressor that copurifies with the polymerase may inhibit its activity on these promoters. Although the primary goal of this study was to purify core RNA polymerase from M. xanthus, the fact that an easily identifiable 6 factor was not present in any fraction during the purification (even those with the highest levels of specific transcribing activity) was quite puzzling. Previous experiments have demonstrated the presence of two proteins (designated 61 and O'II) which copurify with M. xanthus vegetative polymerase activity (73). Additionally, antibodies raised against Bacillus 643 have been shown to recognize 3 proteins (with apparent molecular weights of 86, 80 and 51 kDa) in Myxococcus whole-cell extracts (82). When anti-0’43 antibodies are used to probe M. xanthus core and holoenzyme-containing fractions, 2 proteins present only in holoenzyme fractions are detectable amidst the background (Figure 6) . These proteins appear to migrate at positions corresponding to the lower 2 bands seen by Inouye (82) in her Western analysis. Proteins from both these regions were gel-purified and tested for transcription stimulating activity, but only the 50 kDa region contains a protein which acts with core RNA polymerase to transcribe the Bacillus veg promoter (Figure 7; although the larger 80 kDa 88 protein could possibly also have stimulated transcription if present at higher concentration). Furthermore, a protein migrating to the 50 kDa position of an SDS-polyacrylamide gel is slightly more abundant in holoenzyme fractions than in core fractions (Figure 5). Together, these observations suggest that a 50 kDa protein is the missing o-factor. Evidence seems to indicate that the 50 kDa protein is a degradation product of a larger 6 factor. First, Rudd and Zusman (73) described protease activity as a "substantial problem" in their attempts to purify vegetative M. xanthus polymerase. Of the 2 closely-related proteins (61 and OII) coeluting with polymerase activity, OII was postulated to be a degradation product of 61. Second, bands migrating with the lower 2 of 3 proteins identified by Inouye using Western analysis (82) are detectable in holoenzyme-containing fractions (Figure 5) . Efforts to repeat Inouye's results using whole cell extracts from both DK1622 and DZFl (86) (the strain originally used by Rudd and Zusman [73] and Inouye [82]) have failed to detect the largest o-band. Finally, M. xanthus holoenzyme appears to recognizes the E. coli 0’70 consensus sequence and the 50 kDa protein present in holoenzyme fractions can stimulate E. coli core transcriptional activity. These results suggest that the 50 kDa protein and E. coli 6'70 share structural and functional similarities in the regions necessary for core and promoter binding, which was predicted previously from sequence comparison of the major Myxococcus vegetative 0' 89 factor (sigA) and E. coli (:70 (rpoD) genes (82). Taken together, these observations suggest that the 50 kDa protein and the product of the sigA gene are related. The 50 kDa protein might, however, be an entirely different vegetative o-factor or a 'transcriptional activator. ZMultiple vegetative 0' factors have been identified in S. coelicolor (66) . Unequivocal identification of the 50 kDa protein awaits purification of quantitative amounts for N-terminal sequencing. In conclusion, the major goal of the polymerase study was achieved; namely, to purify core RNA polymerase from M. xanthus for future reconstitution experiments with developmental factors. Identification of developmental 0' factors and transcription regulators will likely require further minimization of protease activity, a maximization of polymerase yields, as well as the availability of a variety of well-characterized, developmentally-regulated promoters. Ultimately, by reconstructing the transcription of signal- dependent genes, the molecular events coupling changes in gene expression to intercellular communication will begin to be elucidated. SUMMARY AND CONCLUSIONS All organisms, regardless of their complexity, need mechanisms to sense and respond to an ever-changing environment. When challenged with nutrient limiting conditions, thousands of seemingly simple myxobacteria cooperatively participate in a developmental program that culminates in the production of environmentally-resistant spores. The multicellular behaviors and social interactions exhibited by M. xanthus during its differentiation can serve as models for higher developmental systems. A complete understanding of the role of cell interactions in the M. xanthus developmental pathway requires knowledge about many components. At the start of the pathway lies the extracellular stimulus whose interactions with the sensory machinery commences differentiation. At the end, rests the morphogenic consequences of a tightly and temporally regulated program of gene expression: intricate fruiting bodies and coat- protected spores. In between, an unknown array of activators, repressors, and modifiers functions to couple the intercellular signalling events and developmental gene activation. The experiments described in this thesis have aimed at the eventual identification of these intracellular mediators. 9O 91 One of at least four signaling events, the transmission of C-signal is required for the progression of development past loose aggregation and is also necessary for proper gene expression. DNA adjacent to C-dependent Tn51ac insertions was cloned to define regions that are important for regulated gene expression. DNA within 1.8 kbp of the Q4414 insertion is incapable of promoting developmental B- galactosidase activity suggesting that additional regulatory regions further upstream may be required for gene expression. In contrast, DNA within 2 kbp of the Q4403 insertion directs developmental B-galactosidase activity. Transcription from the Q4403-associated promoter initiates approximately 380 bp from the insertion point and the levels of transcript are developmentally regulated. This same 2 kbp regulatory region at least partially mediates the C- signal dependence of gene expression. An additional regulatory region lies between 2 and 8.5 kbp upstream of the Tn51ac Q4403 fusion and appears to positively affect the level of B-galactosidase activity. DNA even further than 8.5 kbp from the insertion site may participate in activating gene expression since lacZ fused to the 8.5 kbp Q4403 upstream region produces only 25% of the maximum B- galactosidase activity observed in the original insertion- containing strain. Alternatively, the reduced expression may reflect a position effect (dependent on the chromosomal location of integration) or autorepression by the product of the Q4403-associated gene . 92 Since the Q4403-associated transcript is differentially expressed throughout development, changes in B-galactosidase activity are most likely due to changes in the level of transcript. Although the mechanism of this mRNA accumulation can not be definitively ascertained yet, by analogy to other systems, modulation of transcription initiation is likely to be involved. Core RNA polymerase was purified from vegetative M. xanthus for use in reconstitution experiments with other developmental factors. Core RNA polymerase, followed immediately by holoenzyme, elutes from a double-stranded DNA-cellulose column within a shallow salt gradient. The transcriptional activity of core polymerase on the B. subtilis veg promoter can be stimulated by both E. coli 670 and an approximately 50 kDa M. xanthus protein purified from holoenzyme fractions. The identity of the 50 kDa protein is currently unknown; it may be a degradation product of two previously described o-factors (73, 82) , a new vegetative 0’ factor, or a potent transcriptional activator. M. xanthus holoenzyme appears to recognize [bacterial promoters with E. coli 0'70 consensus sequences. Both the B. subtilis veg and E. coli lacUV5 promoters are efficiently transcribed in vitro; transcription from the M. xanthus vegA promoter or Q4521-associated promoter has not been observed. These latter two promoters differ noticeably from the 0'70 consensus sequence and transcription from them may require additional activators, alternative 6 factors, or the 93 removal of a repressor. By beginning near the end of the signal transduction pathway and working backwards through the machinery, the mechanisms coupling gene expression to intercellular signaling will be revealed. Information obtained by studying M. xanthus cell interactions should be generally applicable to other developmental systems. LI ST OF REFERENCES LIST OF REFERENCES 1. Reichenbach, H. (1984). Myxobacteria: A most peculiar group of social prokaryotes, pp 1-50. In E. Rosenberg (ed.), MYxobacteria: Development and cell interactions, Springer Verlag, New York. 2. Burchard, R. P. (1984). Gliding motility and taxes, pp 139-161. In E. Rosenberg (ed.), Myxobacteria: Development and cell interactions, Springer Verlag, New York. 3. Wireman, J. W. and M. Dworkin. (1975). Morphogenesis and developmental interactions in myxobacteria. Science 189, 516-522. 4. Shimkets, L. J. and M. Dworkin. (1981). Excreted adenosine is a cell density signal for the initiation of fruiting body formation in Myxococcus xanthus. Dev. Biol. 84, 51-60. 5. Kaiser, D. (1984). Regulation of multicellular development in myxobacteria, pp 197-218. In R. Losick and- L. Shapiro (eds.), Microbial development, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 6. Sudo, S. 2., and M. Dworkin. (1969). Resistance of vegetative cells and microcysts of MYxococcus xanthus. J3 Bacteriol. 98, 883-887. 7. Hagen, D., A. Bretscher, and D. Kaiser. (1978). Synergism between morphogenetic mutants of Myxococcus xanthus. Dev. Biol. 64, 284-296. 8. LaRossa, R., J. Kuner, D. Hagen, C. Manoil, and D. Kaiser. (1983). Developmental cell interactions of Myxococcus xanthus: analysis of mutants. J. Bacteriol. 153, 1394-1404. 9. Kuspa, A., L. Kroos, and D. Kaiser. (1986). Intercellular signaling is required for developmental gene expression in Myxococcus xanthus. Dev. Biol. 117, 267-276. 10. Kim, S., and D. Kaiser. (1990). C-factor: A cell-cell signaling protein required for fruiting body morphogenesis of MYxococcus xanthus. Cell 61, 19-26. 94 95 11. Kroos, L., and D. Kaiser. (1984). Construction of Tn51ac, a transposon that fuses lacz expression to exogenous promoters, and its introduction into Myxococcus xanthus. Proc. Natl. Acad. Sci. USA 81, 5816-5820. 12. Kroos, L., A. Kuspa, and D. Kaiser. (1986). A global analysis of developmentally regulated genes in Myxococcus xanthus. Dev. Biol. 117, 252-266. 13. Kroos, L., A. Kuspa, and D. Kaiser. (1990). Defects in fruiting body development caused by Tn51ac insertions in Myxococcus xanthus. J. Bacteriol. 172, 484-487. 14. Kroos, L., and D. Kaiser. (1987). Expression of many developmentally regulated genes in MYxococcus depends on a. sequence of cell interactions. Genes @ Devel. 1, 840-854. 15. Cheng, Y., and D. Kaiser. (1989). dsg, a gene required for cell-cell interactions early in MYxococcus development. J. Bacteriol. 171, 3719-3726. 16. Campos-Ortega, J. A. (1988). Cellular interactions during early neurogenesis of Drosophila melanogaster. Trends Neurosc. 11, 400-404. 17. Thaller, C. and G. Eichele. (1987). Identification and spatial distribution of retinoids in the developing chick limb bud. Nature 327, 625-628. 18. Devreotes, P. (1989). Dictyostelium discoideum: a model system for cell-cell interactions in development. Science 245, 1054-1058. 19. Chen, H.-W., I. M. Keseler, and L. J. Shimkets. (1990). The genome size of Myxococcus xanthus determined by pulse- field gel electrophoresis. J. Bacteriol 172, 4206-4213. 20. Kuner, J. Mg, and D. Kaiser. (1981). Introduction of transposon Tn5 into MYxococcus for analysis of developmental and other nonselectable mutants. Proc. Natl. Acad. Sci. USA 78, 425-429. 21. Shimkets, L. J., R. E. Gill, and D. Kaiser. (1983). Developmental cell interactions in Myxococcus xanthus and the spoC locus. Proc. Natl. Acad. Sci USA 80, 1406-1410. 23. O'Connor, K. A., and D. R. Zusman. (1983). Coliphage P1 mediated transduction of cloned DNA from Escherichia coli to Myxococcus xanthus: use for complementation and recombinational analyses. J. Bacteriol. 155, 317-329. 23. Martin, S., E. Sodergren, T. Masuda, and D. Kaiser. (1978) . Systematic isolation of transducing phages for Myxococcus xanthus. Virology 88, 44. 96 24. Gill, R. E., M. G. Cull, and S. Fly. (1988). Genetic identification and cloning of a. gene- required for developmental cell interactions in Myxococcus xanthus. CL Bacteriol. 170, 5279-5288. 25. Gill, R. E. and M. C. Bornemann. (1988) Identification and characterization of the Myxococcus xanthus bsgA gene product. J. Bacteriol. 170, 5289-5297. 26. Chen, Y., and D. Kaiser. (1989). dsg, a gene required for Myxococcus development, is necessary for cell viability. J. Bacteriol. 171, 3727-3731. 27. Kuspa, A., and D. Kaiser. (1989). Genes required for developmental signaling in Myxococcus xanthus: three asg loci. J. Bacteriol. 171, 2762-2772. 28. Hagen, T., and L. Shimkets. (1990). Nucleotide sequence and transcriptional products of the csg locus of Myxococcus xanthus. J. Bacteriol. 172, 15-23. 29. Kim, S. K., and. D. Kaiser. (1991). C-factor has distinct aggregation and sporulation thresholds during MYxococcus development. J. Bacteriol. 173, 1722-1728. 30. Kim, S., and D. Kaiser. (1990). Cell motility is required for the transmission of C-factor, an intercellular signal that coordinates fruiting body morphogenesis of Myxococcus xanthus. Genes @ Devel. 4, 896-905. 31. Kim, S., and D. Kaiser. (1990). Cell alignment is required in differentiation of MYxococcus xanthus. Science 249, 926-928. 32. Losick, R., and J. Pero. (1981). Cascades of sigma factors. Cell 25, 582-584. 33. Losick, R., and L. Kroos (1989).. Dependence pathways for the expression of genes involved in endospore formation in B. subtilis, pp 223-241. In I. Smith, R. A. Slepecky, and P. Setlow. (eds.), American Society for Microbiology, Washington D. C. 34. Newton, A., N. Ohta, G. Ramakrishnan, D. A. Mullin, and G. Raymond. (1989). Genetic switching in the flagellar gene hierarchy requires both positive and negative regulation of transcription. Proc. Natl. Acad. Sci. USA 86, 6651-6665. 35. Xu, H., A. Dingwall, and L. Shapiro. (1989). Negative transcriptional regulation in the Caulobacter flagellar hierarchy. Proc. Natl. Acad. Sci. USA. 76, 6656-6660. 97 36. Kaiser, D. (1979). Social gliding is correlated with the presence of pili in Myxococcus .xanthus. Proc. Natl Acad. Sci USA 76, 5952-5956. 37. Rosner, J. L. (1972). Formation, induction and curing of bacteriophage P1 lysogens. Virology 48, 679-689. 38. Hanahan, D. (1983). Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166, 557. 39. Messing, J. (1979). A. multipurpose cloning system based on the single-stranded DNA bacteriophage M13. Recombinant DNA bulletin, NIH Publication No 71-99; 2, No 2, 43-48. 40. Meissner, P. S., W. P. Sisk, and M. L. Berman. (1987). Bacteriophage A cloning system for the construction of directional cDNA libraries. Proc. Natl. Acad. Sci. USA. 84, 4171. 41. Id” 8. and IL. Shimkets. (1988). Site-specific integration and expression of a developmental promoter in MYxococcus xanthus. J. Bacteriol. 170, 5552-5556. 42. Gill, R. E. Personal communication. 43. Kroos, L. Unpublished data. DK5208 was constructed by growing Mx8clp2 (23) on the csgA donor strain L8523 (csgA- 205 Tcr [44]) and transducing DK1622 to tetracycline resistance. 44. Shimkets, L. J., and S. J. Asher. (1987). Use of recombination techniques to examine the structure of the csg locus of Myxococcus xanthus. Mol. Gen. Genet. 211, 63-71. 45. Maniatis, T., E. F. Fritsch, and J. Sambrook. (1989). Molecular cloning; a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 46. Stellwag, E., J. M. Fink, and J. Zissler. (1985). Physical characterization of the genome of the Myxococcus xanthus bacteriophage Mx8. Mol. Gen. Genet. 199, 123-132. 47. Sternberg, N., D. Tiemeier, and L. Enquist. (1977). In vitro packaging of a A Dam vector containing EcoRll DNA fragments of Escherichia coli and phage P1. Gene 1, 255-280. 48. Avery, L., and D. Kaiser. (1983). In situ transposon replacement and isolation of a spontaneous tandem genetic duplication. Mol. Gen. Genet. 191, 99-109. 49. Miller, J. (1972). Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 98 50. Bradford, M. (1976). A rapid and sensitive method for quantization of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 73, 248-254. 51. Igo, M. M., and R. Losick. (1986). Regulation of a promoter that is utilized by minor forms of RNA polymerase holoenzyme in Bacillus subtilis. J. Mol. Biol. 191, 615- 624. 52. Burton, 2. F., C. A. Gross, K. K. Watanabe, and R. R. Burgess. (1983). The operon that encodes the sigma subunit of RNA polymerase also encodes ribosomal protein 821 and DNA primase in E. coli K12. Cell 32, 335-349. 53. Gilman, M. Z. and M. J. Chamberlin. (1983). Developmental and genetic regulation of Bacillus subtilis genes transcribed by st-RNA polymerase. Cell 35, 285-293. 54. White, J. Unpublished data. 55. Romeo, J. M., and D. R. Zusman. (1991). Transcription of the myxobacterial hemagglutinin gene is mediated by a 654-like promoter and a cis-acting upstream regulatory region of DNA. J. Bacteriol. 173, 2969-2976. 56. Downard, J., S.-H. Kim, and K.-S. Kil. (1988). Localization of the cis-acting regulatory DNA sequences of the Myxococcus xanthus tps and ops genes. J. Bacteriol. 170, 4931-4938. 57. Kil, K.-S., G. Brown, and J. Downard. (1990). A segment of Myxococcus ops DNA functions as an upstream activation site for tps gene transcription. J. Bacteriol. 172, 3081-3088. 58. Apelian, D., and S. Inouye. (1990). Development- specific o-factor essential for late-stage differentiation of MYxococcus xanthus. Gene @ Devel. 4, 1396-1403. 59. Hama, C., S. Ali, and T. B. Kornberg. (1990). Region specific recombination and expression are directed by portions of the Drosophila engrailed promoter. Gene @ Devel. 4, 1079-1093. 60. Banerjii, J., S. Rusconi, and W. Schaffner. (1981). Expression of a B-globin gene is enhanced by remote SV40 DNA sequences. Cell 27, 299-308. 99 61. Moreau, P., B. Wasylyk, R. Everett, M. P. Gaub, and P. Chambon. (1981). The SV40 72 base pair repeat has a striking effect on gene expression both in SV40 and other chimeric recombinants. Nuc. Acids Res. 9, 6047-6067. 62. Lewis, E. B. (1950). Advan. Genet. 3, 73-115. 63. Feinstein, S. C., S. R. Ross, and K. R. Yamamoto. (1982). Chromosomal position effects determine transcriptional potential of integrated mammary tumor virus DNA. J. Mol. Bio. 156, 549-565. 64. Hunt, T. P. and B. Magasanik. (1985) Transcription of glnA by purified E. coli components: core RNA polymerase and the products of glnF, glnG, and glnL. Proc. Natl. Acad. Sci. USA. 82 , 8453-8457 . 65. Grossman, A. D., J. W. Erickson, and C. A. Gross. (1984) The htpR gene product of E. coli is a 6 factor for heat shock promoters. Cell 38, 383-390. 66. Buttner, M. J. (1988). RNA polymerase heterogeneity in Streptomyces coelicolor A3(2). Mol. Micro. 3, 1653-1659. 67. Chater, K. F., C. J. Bruton, K. A. Plaskitt, M. J. Buttner, C. Mendez, and J. D. Helman. (1989). The developmental fate of S. coelicolor hyphae depends upon a gene product homologous with the motility 0' factor of B. subtilis. Cell 59, 133-143. . 68. Inouye, M., S. Inouye, and D. R. Zusman. (1979). Gene expression during development of Myxococcus xanthus: pattern of protein synthesis. Dev. Biol. 68, 579-591. 69. Komano, T., T. Franceschini, and S. Inouye. (1987). Identification of a vegetative promoter in Myxococcus xanthus: a protein that has homology to histones. J. Mol. Bio. 196, 517-524. 70. Downard, J. S. (1987). Identification of the RNA products of the ops gene of Myxococcus xanthus and mapping of ops and tps RNAs. J. Bacteriol. 169, 1522-1528. 71. Downard, J. S., D. Kupfer, and D. R. Zusman. (1984). Gene expression during development of Myxococcus xanthus: analysis of the genes for protein S. J. Mol. Bio. 175, 469-492. 72. Beynon, J ., M. Cannon, V. Buchanan-Wollaston, and F. Cannon. (1985). The nif promoters of Klebsiella pneumoniae have a characteristic primary structure. Cell 34, 665-671. 100 73. Rudd, K., and D. Zusman. (1982). RNA polymerase of Myxococcus xanthus: purification and selective transcription in vitro with bacteriophage templates. J. Bacteriol. 151, 89-105. 74. Burgess, F., and J. Jendrisak. (1975). A procedure for the rapid large-scale purification of Escherichia coli DNA-dependent RNA polymerase involving polymin P precipitation and DNA-cellulose chromatography. Biochemistry 14, 4634-4638. 75. Jendrisak, J. J. and R. R. Burgess. (1975). A new method for the large scale purification of wheat germ DNA- dependent RNA polymerase II. Biochemistry 14, 4639-4645. 76. Kroos, L., B. Kunkel, and R. Losick. (1989). Switch protein alters specificity of RNA polymerase containing a compartment-specific ofactor. Science 243, 526-529. 77. Hager, D., and R. Burgess. (1980). Elution of proteins from sodium dodecyl sulfate-polyacrylamide gels, removal of sodium dodecyl sulfate, and renaturation of enzymatic activity: results with sigma subunit of RNA polymerase, wheat germ DNA, topoisomerase, and other enzymes. Anal. Biochem. 109, 76-86. 78. Matsudaira, P. (1987). Sequence from picomole quantities of proteins electroblotted onto a polyvinylidene difluoride membrane. J. Biol. Chem. 262, 10035-10038. 79. Moran, C., N. Lang, S. LeGrice, G. Lee, M. Stephens, A. Sonenshein, J. Pero, and R. Losick. (1982). Nucleotide sequences that signal the initiation of transcription and translation in Bacillus subtilis. Mol. Gen. Genet. 186, 339-346. 80. Silverstone, A., R. Arditi, and B. Magasanik. (1970). Catabolite-insensitive repression of lac promoter mutants. Proc. Natl. Sci. Acad. USA. 66, 773-779. 81. Kaplan, H. B. and D. Kaiser. Personal communication. 82. Inouye, S. (1990). Cloning and DNA sequence of the gene coding for the major ofactor from Myxococcus xanthus. J. Bacteriol. 172, 80-85. 83. Burgess, R. R., A. A. Travers, J. J. Dunn, and E. K. F. Bautz. (1969). Factor' stimulating’ transcription. by’ RNA polymerase. Nature 221, 43-46. 101 84. Lowe, P. A., D. A. Hager, R. R. Burgess. (1979). Purification and properties of the 0' subunit of Escherichia coli DNA-dependent RNA polymerase. Biochemistry 18, 1344- 1352. 85. Morrison, C. E., and D. R. Zusman. (1979). Myxococcus xanthus mutants with temperature-sensitive, stage specific defects: evidence for independent pathways in development. J. Bacteriol. 140, 1036-1042. "Iillillliiiiiill"i'iiiiii