..VVVVVVVVV. mgdwwkfi...” V..
%Q.V.V\VV\\2¢1, :3. V V V
skis 143.1.»ch V”... N. v ..V ..
2.49.? 11:11 .VVV. .
0.145471th 171 1 1"}: w 4 .251)...
J! «”71. V V...

. V V
x .WW.VVVV\V$N u :«M. .3 Via}...

a. a... me (n11k..\x.3r4 (9. 1 V
.

   
    

       
    

..V: VV‘VL, V .
a... mafia .
\..‘..x, V
V. V . -

           

 

V. ,. , , .
.. .. V . . V V V V
a. «I v . l bl I .
Nmmuu» 1.5.31». VVNPL)? . v\ , 1

. 170,. . 1 ,. :1 5V“... \ V .V I
, l, VanwuVuflumubwu V V. . V. .1 2.x». iVVVVVdeVUL, VVV.
.. V. .V V. V. ..V-\VVvuhxinfl 1.. . 3V V. V - ..V. V 206.“) gamma“): ,1 V 10.09 V. A
V... .V. 2V¢V_SV,VV..s%?V..?\\...,V\.mm.VVxV ..VVVVV huuKVfla VV V V 3. 5b.: .1st: V. V .1 1 .V_ V
7 RxVVVEVJV. .15.: 35...... 31,53. 1...? .. 2V4; . .v VVVVW¥§VOQV2¥15VV V15 151
.3qu .. 5.1.1.2.... his; .Quuu...V...VhsxV.,V..VV.:VV¢VVVVV.V:V . . . .. V ..V): hyflrflghuxfi1§VV3L3L13thhs v.1
i . V V HS..V{V.th\....aV. 1» 1 . .V . ,V .. . . .V $1.5 gxirfir V11 “HW‘
. VVV 5 VV ) V. V V n. I 5 V.
. .. . . 3.2.3.1
. V ..

bk sauna.
Emu III
.V)\\\§Vl$u.fl V V V V V
v .5115 \‘INTIVVIVJ‘ VI Luv‘ltfluu
1’11
111V
:1

         

       

 

V
(1 3
an H...

0.1.1518\1\V1~1 ' 1
i x

is , V V
fabuxpgxfihififix

 

V V
V V 91.111.111.111.
)1... £111 ’71:: i V. . V V n. 1574111 .
. ..eruVuRmmV “1.20354 V ..V... V. . 1h fmummmewasaflizdnfl?
i. 1 {‘0‘ ..er 1 ‘V 1‘ V VS“\ 15 v..
”6.111! 13V.V£N‘|r| “"111 1b; 1V\V:A V1. fi))x)$1Vw.u8111\xTV 1a,:- IV. hVIVi‘QV‘VVVw1113V ”63»va Vl. . V.
. . V . .73....H3ig 11.21....VVVVVVV... V... 1-3.11 3V3... .15.... 3mm ..VIVVVVVVVI 3... till. .
. V V. at? \V.V..V.,?VV. .cmansV.x.V.V..u.V.qVV $1.... ...Vs..V..VVVVsV.VVVV .. Max:113)?..8..Afifihhshhyvi%hax§{ fig). VVVVV V. V p.
aim“... 2.12:3)..1dénx timiflaniV ..VIKVVVWVVVVVIVVH “VflsEu.nva§xhax \Vtsxxntsrxx. Nu...) l 15.23... V...
V. .51... vi Vii]... V Mail... . 5h

. . 1 I. nu. .. \1)\().:\ 3151.... 3115‘.‘ . .b\ ".71.!
\ IA . . 2 l i '1 .4 V 10‘ V. V: s V
4Vsm§ .V 3.14% VmHVVVVfiLsQVWVV .52}thth waVuLleVkVVVv}Valiausivnwwt.
V. 1...» z». . .. V... ,. 3b .

,. ,«VVV uhnfishwmfi ..stVl},..;53..VVVV.%VVm.VVV.8V.V Rhianna”... V «a?
. V» V

10...}ka 1.937151 1.5843 771V T1234

11: 101111 JV. .3 \.w x
1...: V a V. ..V .1 V V . V guns? 733113 V . V
V .. s. V 4... V. V... V. .
agsflphwasfihfi .VmVnVVVVSREmLVdebV wingiflw: £23.... 3. V .1“...st V...“
I V V :5 a . V 4 ..V. V V n 5 111+. \\\z.wflm.\ksv..\v1 1...!

V. V. )5 m \D 2))? V ‘51)}
. Vic .VwmwwuugwuflflhughfiV V. V ..V..VVV.M,VmVVV.VVh§mLW $9.0qu ..V... Zayfiiflfikk 2”»??st 2 Quin... 1.1% n
V. ,V. 1. V. kin.) .. In . .V. ..1. V. Ln. . . ..VV . . 3 ,V r. a... V V. .
I .61 V , B mmnszfiotmu. V 0911 11,,qu 2., . V 1 .5 ..Lu s 1? 2.. V. 111 TV . . .
V . V . V a .

:
it...“ .13.! fugigVVthxz u \I. an 3‘
\V

           

   

   

 

  

        

 

                                                  
                 

     

1.1.311‘1311 I
. V 551:1) {13)}...1 ii
' a V) L A12.“ all 1 I 5“ \ ‘Qs
V( BEG? V. .3. 3 V}. 3. V).r.V.\n\va 31.33.3731 VuVVVMVVis .1 V0}: V
1. $41, 1. 1 It 1&1. I). .1.) 113.11: . 2:; 111V. J51 1...... 11.1351Vittllluk I? ‘VJ-‘JGI! 1:
V . gr.“ \Mv 31.5.1351 1.4\ .VVV.\1V.?‘ : 33.31%}... .V. . .\ VJ): 3211\51811111315 1 $32.11!]. 131 ..‘.I\\.1.vVIV’\\
. s V. . V a. (.3: i452}. «1.1.14.1 1\31.4)$H\.Vl.1r , ‘1 V “171% 322 1.1. .1§513§V.11113V%
13151.1(- 3!‘ (HQ JV) 1111,11). )1 .11.. 2311:. ..V ..V. V V . V V1.11 V43. 5115): 37:§3}:.-..r1321
2:53.. .3... . .3325 52 VVVCVVVV.V€\V5 \. 11.2.. ..V V. 11 .V .VVKQIQVVigVIIziiVVJS 1.2. .
V 3112: VanV V VF“. Why“): V. ..R.)Jwafi..u»2¢5.1.. V. V V. V. V. V V.,V>\ 12!...» : ,VV.11.{_\.VuVVflcYV<1\}J. .. {1:50.11 V .4: V
v. 3 \l i.‘{ V A V. , \0‘;;‘. V‘- 1 I V . Wm) z
TV’VAWU' 1311 ““31”; a ’1 \7 ‘11].:K11!v{wulnfluu.1713‘\$. V. . 3313;11315’51335 1111 131. 5.1111131 .
g 3%} V) 1 . .VVIV‘VQV. 1.111531: .4 V... 1. 15133112111511 10537.11. 1ng 1.1313: giiizlgpsvi)‘
1‘. V , V . \ V. V .V \ggxivinvnt V V. . . ..1..11.VI...(.\..VV.1J»VV. V‘QSKVVVJVQJV ..\\.L.J J19>3\VV)3\1 V V 111
. 1.75.31 13.1 V V ..V 11.....(15123Vfi53 . . 1.3.2.2135 1.1.1.11 125.3112.» . .5 V .V.1 {N311}?
1:131:11 . V . 4 .V V I V 1‘. 32113115 1.; 2112)}..V1V1x1h \ 113.11 7171\1...‘ \\.V XVJVVV ..Vié‘Q
V V {\HiVllachJV‘ 114 :9). 11...} (a\\. F1151} 115%) )fi’lafig 11.11.... ‘Vu $1.070)...ng
1 13111.33...» V1311‘1V‘I 1.11»! .1 I)? 3] 39.1 193)} 1.
. Viv-Via?) )1“... L7 V 3511 1.3)}11. . 11:“! 8115133211111 .
(ft t. V 2; 1.- 3111\1 \ {111.1111 3 Vr)‘ V 115331411... 33.5111. 1 ix}... «(13‘ 3x... 11h]:
V V t... . . V v V 111.33}..qu 37511! ‘14)) 1111. .1\V.\.V.3I 41>VVAJxfi5u1V. J. V1 a J\ 7‘\ 3V. 1». 111.1111)! V.\o\l.1 341,13.
9151!! .175“.va ‘r’l. VV . , V . 1.x MVH137§MH>JV5IIV11VVVVin4l 11371.1 (.11.:5ws). 7 k V. E}J§.u1xfi%flf$1¥flinfiyflfi 1‘ J: l 11

ID. V II!!! , . V . 11 V glii! IVJSV. \1 ,..n .111 . V. {1313131 V33.“ 1 V

1.! .5in 1% ”'90.: V V V . . V . .V v .V 5VVVVV055VAV3J1VV11)(.\ 113.3.Vf11§5§$\xfi( \{Vuflni 13h1?‘_v.(.131\311131)\1 13111552.“.3.
¢ 29'§;§&l‘f 09-. VII .. £11311}? 1)Ll.)\1.1.).11.1131113VVV15911¢1\.)21:.191111u1s14721VJI31V13V)!’ V 41.11125111’121
lVV.v§v6!)bliOI-§: .1‘ , éVv Ely?! V (kW-VIVVVIJJI. 1.1 l.}(.}\{? 134411315. )1:...111.1)7y\51.1~.»\ {11-1137mfl1i .levaajxgfi
A r ..u. . AVIVT. .let .f 537.1.” Iii 5": V \fi. 41.1.1.1 54.8521 72:113.“.0213 \s‘o?!\«u11\3}1‘\N\VVd-\‘ 21“} 11‘ 2})ii12’. )\

V .. VVu 1‘11"» 1. vuugr . . V \11‘3.111\\J¢V;31111 .5312; \ \Vtiln‘! 1.“. .4331 13.1. 1V(‘\ifii\:\\
1 n s.

   

   

\l‘.
1V1“. 1 1.3%. VfilZnVVllfilw. , sllu 1.75s. 1") .
than... .12.)... .3393? 12643.53.than VVV....Vh.Vn..VV.VVVVs..V.xw.\2mqumVfiLV
Ultk».$\»\131\13.}1 15,. W‘firlj. 2‘) Au“

.. V . . 1153.. V
.V 7)) 1V1. 1 (1712555311. 11}!:S.1>1V11111111¢QUV\){..V1 .111 It}
11)..“ 1.11 \V.\Sl....1\111 3.115% V. 17.1 1142.}: #3751313? hi

  
   

   
 

  

 

   

     
 

           

   

\. V \
I \ | 1! V V . . J V ., V v V l \. V t! . V V
111131154 ..Vrhéc.‘ 11.2)...in .1 V. .111 H1111: . .15. . . .V .
\gu...‘1‘12.\3110“i\3\nrhfiul.q;m\tflv1flhvfflh¥thwseryfiaeru§n In”! 7} V E 1111.! \V, ._
{It 511.]. 74.1.1111.) $311121V1l1t 1741.1.311513I5‘fiw41‘173ifiz a! .xflkuflfii \\ V V ‘0
Vx143!V511¥511$).11‘\\1.VV.\V)15.(14.151313)V\1...:1‘. Vic: 313111.31 Ila! V. V g u t
131115113. 11.3 13.1 )1! 1!.1iuetfxu131 1111 2.113311111135:541vvfi\\ri))1 V. 2!...) Va)! 1 . ‘1.
(1)113V311V11de V.)VV(,1hu¢VUVV.3~) V l]....1\\$\7,:h.15)\x114111.)15\1\(<V$)}\))5V.\ $11.! 5 ggTYhRVI
V 1’ V 2:. 4 (5.6.! L.“Ht£i\l|1u.‘r_ IiBiLV 11.1.13}! §$hflVcV§i§5E§§ 1.11 V
. V (lift? ,V V I. .V ..V V \V .V a. 11’?!le C. ‘KIlV‘V-zzimflvviiVI )1)“ r3191)?! I 1:131; \11115221?dbfl.ldh)171\ %Ni .9? I
VVVIVVIQECQRIIE'I «1| )011 . .1 . . . V 111V ~11.\nl;721131§1)1\1) 513111-1111! J1\\ 1.1.1. 4313111311..)131. 1. 1%)}: 521)}!
Xvi?! 1190 V V J 1 1%.éo!+ 3111:1113)! (19%}; Viki 115.413; V. Vifiifi i.
4! ill‘ii‘ r11)“; V V1,] )(lN. .15.)!7111535. 1.3!! V‘TVaVIVvazVIVIVWJVLVVvt , 1.1.1... V.v!3..!).o}l.. SVVWXEVVVV £31 w. ’11:...
V 1%{59-5! -.V fl: {’é‘111wgdr1/lti 1.11. fflust 4th,?V:Rangilg;qugfiqV~§§&3)a1éh% 1 V q. uqu 1
E. V n V. .1 In) 4. 11.- VVVIV‘ V... 1.1.1 3.31335121211 .V... 151113-51 . 1V- I
fir-EV. '11:... V 11‘VQVVL331...“ «igfiqzngmixliaiirgaaflélfll511hv21321415) 3.55.14. , gxlhwmfli Z. a: .1. 1
05...! 31.1.1 V%\ 134.11111VV1V ringixsxgwhix Inlivfifinrirsauui;ku1iuw‘¢uv\xxl. 5 111.35} «11 2
. I1 731% 7.111 .311... 3"... V IVE-172.1»?! , V 3.119110!
11.06. IQ!" 1.1:. .l... 11!. 1 9J8: . 1.1.1.7 15111111591119.1311\ g¥5§19111311§1 1(113231’1V§\ xii-1‘17 £5.01... ‘01.
if: ‘32.???11 . 1.41 11111114111}; 11151131131. .5 .12.... JV)! 1.. 1, \K‘ufl‘tfixl .131. , ’\ £11!-
OVIVIOD V $4134: . null... )5 (.51)! VI. VJ! V¢)>\\.|v 215211711 3 )‘Vt‘VVnV. ..l‘ 4.9V 1.1 ..V. I 1.310%“. V ‘1‘ i Vi? . V
I! 53... 5.14.1 gfiV‘iklhnnlxgHuuig l:110fl¥311}fl.§11§fl!§\§fl1 figlgfiigflflvh. ..IVJH} {Sufi}!!! 6L
. If \ 711.1(- . 1’11! in V: I. 51, 1:!) 4,21... 1.111 I f;
'41!er? V . :27? V V :V 93'! V. .erV 1‘ lhifluhuru‘f “W’shwgllni1uvsz11fifinlfiwu. @“raVfifin (11 4104151 {g1oflfla‘ufi'1 VI!!! ‘1 b, . 13%". V
.rf E10131}! glVéoIfi-‘zgfi E151..V1n§.\1 s an. 4.1.1111: 5.81.. ’:\!V.V...V\%3V1VV .hvl‘...xu V. ,mVIVVV‘VeVrfii. Thu“... fiv15LH1VV... V “un“) fig 1.. . .\I .01 VI 1. .
1.9.11.7!!!» Cirillo-hurt! I! in... II 1. V «1!! 1 11": . ‘1).1 1,331.3.1‘1" 1.: 133.51%... )Lgr)l%$t\wih 1..“ 314.511; 1r 1. V. I‘m): 3113‘. (”“3111 I V.
V V... 5.11;; 11.1 x)...- 31}?C‘é313:§1.l¢ 5.«v1.51A))fll.\.§..€ll.11flllll$(ti.§, 97,1113 I «3(1) 5!
’ffhi 1:th . V \\\|\Il11\i)|\.\.\g{v)71 11s! v..V\.IS.I-J!§)1)\.§V v.1 11- 2:. 9)); {1111? 32111111‘3151 , V; V to... $1 .11.»
. $71.11.. 1))2‘ )1“ 313.1): ind-1112.“. valisiVi‘libfijfiE. 71.;‘515Vfiu131 , 41¢. .
. 1:113“ , 1(1111. V1311) \hti: 31.1 «f11131‘131gzggl11! V
25111511. . $511611 1.115;}? 33““ $2.: lgixlig V1112? ..
. 3241385 V 3 V 1! £1? 5511.5..1‘S‘hv V 131411.}?35! igii‘. :
.l 571%» V 11..;\1u\§§vll¢qi 19.1! V17. .1“. $111.1‘i \V \ V
)3?qu 1;; V31L‘ht‘.\\.\6llii\1\l\’1 7191 3‘1. «.17.. :1 V , L2»!
.11! 15311211..» 1113.?! .5313}! 3.11 1.1.4.11 VV.V inn «, .u .131.“ 1 ,
.. g?! V 1. $1§1§3fi11311 “NV... 5"...» 1:113333115 . (9&1 b.1108...”
uh . . ‘13) N. figilliznifl. w L. ..V. 1111-2311.. 1.11.61 As 1 .
V. 1'13, VII-iiilr‘hvxl x twp-41111.15“? . 1) g . V nV.| V 1\V3~%.(.:IVJ.811_VUVIV $1.11. V}; V3h.>_V.V..€Vhi1...VAWVVmulu..V V. . 31V .111 V . 3 .9115): V1 \
Vi % 1‘21... 1’11): 11.31211: 3153.1 ~ 1311!? . 5.131013
1 )n} V .4131 in" )1 1m! $11§§10H41§y1 “V1131! “11%“?1511352 1% VG. V1.6;
v it ul1w| $1.»; 1’ V \I 141! g .
V7131. 151.55.517lflfit5fi 12$iq§04|§311¥15i111i1 1911221,}.VV.%§~§§ . 1%
1 . V: V £2.13??? (41!: V . 313.13.. 531.15) 1 5.5V
V . 1111.13.43 211.3321 7?!\)\g¢.}1},§ 151.1% V 71. V
1513.5.é3qliiwgi V \ V I}! V .
.327 1.1.}. V
ll {1.11.111}!
1}}HIVII‘V V3.11).

            

V .
275? \ 1591.11.13 . V... . V
staginz (VI-Iii. V .11 it!“ 1:: $10.». ,7. LGILQLI NV. “WM...”
V . u?. I 34%;; , V II I?
V >5. 2.131.... .5? ..V...- V an“. gsfiuflduvi t... a).
t: . .1

3. 31' V $0110. it}; VLV|A,Vi\I.
V V 2. L giglfhxi trait-i...
. 1.316. Vi 53.5.1 1. 01‘ bit
fin“! u {mmgigllIltnkthxa

V. 1.:11 \Vf..xl..A.VLrVVV V 1.5..
.Vtrlhhhiltigfzgh

      
  
  

     

V it ..V. .
. . . V V «Jerabuixmr
V ..V..l..VV.:... V V. , 22H“. 1 . , V V. ..ttfst! V.
V «.V .VrEVtuDk‘; .VVVVV l. .I.. iltoulifgttk tit... V . 1 . . I
rfudu!§!.lhflfuntfifu.i§. .117? ..V... .1! . VVrVVVrVV £1? V . . ..V V --V!
11 V \ VF» IIVI‘VV nll.‘ . k If. in V V
V t;! r. its“... skugflqlflti V. .VVVV V... V; . V VV. V. ,V

 

V7131: {Iii/t; 5.371;! 41. V1.9;

 

fun $331

7 tr 5‘11!
{tidiéf V A

             

  

- n V 5.!!th
VVVVITVI |VV.PV.V1.L V...
I1I§AVVALVIVR . g”? i 5.1"
(V .h in V tVVV-Vlofif’kr I. ‘Orlluntzo"; nirahtlltl: in.» £97.
V VV V V 1:11Iifi9. .85 initdb‘hhflbhfk . 1
V ngfl‘tfinurria if .,.v\? ~ _ V

.. Vt...
. (.1...th 3;..311xii.§l:..§xinflrlf ..V
l.1:57....VIV11119V1Villtxl..- k... . ..V. 3.2%..
V . . V4-1.63.1:{VVttis’thtVtleLILbri«zitgiit... V
V V 2...; . V. .

 

 

              

    

 

     

    

 

  

..V
V
L 1) V V. .V. . .. V... VVV V VI- .L
b VIII. )V'AIJVxIAvftr "113.1 .LV.V.\ lxh..fl)1}~NL/.¢ .rfiu‘VvaVV-Vi1l1r11viitt‘|. «fikaVerttleVx V 17?.) .V .31 fri.‘ villi}! ,boVIVL l1 L7rx
VV .V .V V . V V: V. .VV. :1 V 1-1+? it: .311! .3!) 1V, .erVV . . 1.... .V.... V V.... I.-. . .V ..V)...
V V V V V£¢V‘V. |.\..§?.S.V V .\.V .4; to... {Sufiliéfchlfirkflusv‘fiznrtfVVVVrV 1.5.: .1; If: VVLVVVV 1 V V.I.VVV..\. F111:ttl;t.,..»fi.it\l\$.l
.1311”. (1“. -.VAVinJvaWVFIVrfi, VVVJVvVuuVVVRV... 5.5..51...»V..VV...$V¢. V1.VcVV.tAVx.vK.f»V.V.V§.V £2.11. 1.3:... E. .x. V V V. ,VVVKrpwVVJrfiVE11VuthV V1.2: VV 14V

r VV V V . VV rV>VV

 

 

This is to certify that the

dissertation entitled

ILLUMINATION OF FAT GLOBULE CLUSTERING
IN CONS' MILK

presented by

John R. Euber

has been accepted towards fulfillment
of the requirements for

Ph.D. degreein Food Science

 

 

    

Major professor

Date (1‘19' 829-

MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771

IIHLIHHIIMINIQILIJWIIIIIIIWJMIHII

 

RETURNING MATERIALS:
IV‘ESI.) Place in book drop to

remove this checkout from
LIBRARIES your record. FINES will

 

 

 

 

 

be charged if book is
returned after the date
stamped below.

a? 3%

JUL 2 4 1992‘

e53

 

 

ILLUMINATION OF FAT GLOBULE CLUSTERING
IN COWS' MILK

By
John R. Euber

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Food Science and Human Nutrition

1982

G // 73/00

ABSTRACT

ILLUMINATION OF FAT GLOBULE CLUSTERING
IN CONS' MILK

By
John R. Euber

Fat globule clustering is a prerequisite for normal
creaming of milk. Fat globule clustering as characterized
by the cream volume and cluster time was studied in raw
milk, heated milk, and in model systems. The influence of
various milk fractions, specific anti-sera, and simple
carbohydrates were evaluated. Immunoglobulin M (IgM) was
confirmed as the heat—labile component involved in fat
globule clustering. Its participation was also demonstrated
using raw milk. IgM was shown to function as a cryoagglu-
tinin rather than as a cryoglobulin as previously indi-
cated. Hapten inhibition studies demonstrated that the
antigen is carbohydrate in nature. Skim milk membrane
(SMM), isolated by ultracentrifugation or salt fractiona-
tion and gel filtration, was identified as the homogeniza-
tion-labile component. Although IgM can agglutinate milk
fat globules (MFG) to a limited extent, normal creaming
requires both components. Approximately 7% of the IgM in
normal milk participates in a single creaming. The lower
portion of creamed milk (skim) failed to support cream—

ing upon addition of washed MFG, but did so upon the

John R. Euber
addition of SMM. The presence of SMM in the lower portion
indicated that not all SMM is capable of participating in
creaming. A theory of fat globule clustering consistent
with the observed experimental results depicts IgM inter-
acting in an antigen-antibody mode simultaneously with SMM

and MFG through specific carbohydrate moieties.

ACKNOWLEDGMENTS

I would like to express my gratitude and appreciation
to my major professor, Dr. J.R. Brunner, for numerous
discussions during the course of my graduate studies and
his help in the preparation of this dissertation. The
influence of his investigative approach, encouragement,
and patience significantly contributed to the success of
this study. His personal philosophies will continue to
serve as a source of inspiration.

Appreciation is also extended to Dr. R.C. Chandan
and Dr. R.C. Nicholas of the Department of Food Science and
Human Nutrition, to Dr. F.H. Horne of the Department of
Chemistry, and to Dr. H.A. Lillivek of the Department of
Biochemistry for having served on my graduate committee.

The assistance of Ms. Koch and several fellow
graduate students during the course of this study is also
gratefully appreciated. I would like to thank Dr. C.M.
Stine for the use of equipment during this study.

I would also like to thank the Department of Food
Science and Human Nutrition, Dairy Research, Inc., the
MSU Graduate School, and the Dairy Remembrance Fund for
financial support provided during the duration of my

graduate studies.

ii

Finally, I would like to thank my parents whose
encouragement and support have made my educational goals

obtainable.

TABLE OF CONTENTS

LIST OF TABLES.
LIST OF FIGURES
INTRODUCTION.
LITERATURE REVIEW

The Role of Fat Globule Clustering in Creaming.
Components Involved in Fat Globule Clustering
Physical and Chemical Factors Influencing
Creaming.
Low Temperature
High Temperature.
Homogenization.
pH.
Added Salts
Creaming in Milks Other than Cows'
Mechanisms of Fat Globule Clustering.
Facets of Fat Globule Clustering Narranting
Further Investigation

EXPERIMENTAL.

Materials and Equipment
Preparative Procedures.
Immunoglobulin M. .
Immunoglobulin M Fab Fragments. . .
Soluble and Insoluble Protein Fractions of.
the Milk Fat Globule Membrane
Skim Milk Membrane. .
Preparation of skim milk membrane via
preparative ultracentrifugation
Preparation of skim milk membrane via
salt fractionation and gel filtration
Milk Fat Globule Membrane Gangliosides.
Casein Fractions. . .
Preparation of whole. casein
Preparation of K- casein

iv

 

Page

.viii

xi

moo

._l—_I_J.__J__l—l._l——l
\ICWU'IUW-bN—'-'

NM
w—J

NNNN
00-b-boo

woo
NC)

0.)
N

0000000000
(Jab-boom

Preparation of caseinomacropeptide.
Chemical Procedures
Lowry Determination of Protein.
Babcock Determination of Fat.
Electrophoretic Procedures. .
Discontinuous Polyacrylamide Gel Electro-
phoresis with an Acid Buffer System .
Discontinuous Polyacrylamide Gel Electro-
phoresis with an Alkaline Urea-Containing
Buffer System . .
Sodium Dodecyl Sulfate Polyacrylamide Gel
Electrophoresis . .
Immunological Procedures. . .
Two- Dimensional Double Diffusion.
Two- Dimensional Single Radial Diffusion
Creaming- Studies Associated Procedures.
Preparation of Samples.
Milk. . .
Heated milk
Homogenized milk. .
Washed milk fat globules.
Synthetic milk fat globules
Quantitation of Creaming Capacity
Cream volume. . . . . .
Cluster time. .
Enzymic Treatment of Milk Fat Globules.
Spectrophotometric Evaluation of IgM Cold-
Induced Aggregation . . . .

RESULTS AND DISCUSSION.

Isolation and Identification of Immunoglobulin M.
Isolation . .
Electrophoretic Analysis. . .
Two- Dimensional Double Diffusion. .
Confirmation of Immunoglobulin M as a Creaming-
Active Component. .
Effect on Creaming of Heated Milk
Identification of Immunoglobulin M as a
Creaming-Active Component in Raw Milk
Removal from the reaction with specific
anti—sera

Removal from the reaction with 2- mercap—:
toethanol .

Effect of adding immunoglobulin M to. raw
milk. . . . .

Immunoglobulin M Cryoaggregation.
Quantitation of Immunoglobulin M Participating in
Creaming.

Page
36

37
37
37

37

Examination of the Ability of Gravity-Separated
Skim Milk to Support Creaming. .
Creaming of Recombined Milk- Washed Milk Fat
Globules and Gravity- Separated Skim Milk
Creaming of Recombined Milk-Washed Milk Fat
Globules and Gravity-Separated Skim Milk
from Raw Milk Supplemented with Immuno-
globulin M . .
Effect of Immunoglobulin M Isolated from
Gravity- Separated Skim Milk on Creaming
of Heated Milk . . .
Examination of the Role of the Antigen- Binding
Pr0perties of Immunoglobulin M in Creaming
Isolation of Fab Fragments
The Effect of Fab Fragments on Creaming.
Creaming in Simulated Milk Containing Synthetic
Milk Fat Globules. . .
The Effect of Soluble Milk Fat Globule Membrane
Proteins on Creaming . .
The Effect of Milk Fat Globule Membrane Specific
Anti—sera on Creaming of Heated Milk
The Effect of Simple Sugars on Creaming. . .
Fat Globule Clustering Interpreted in Light of
Immunologic and Serologic Knowledge. . .
The Effect of Enzymic Treatment of Milk Fat
Globules on Creaming . .
The Effect of Milk Fat Membrane Gangliosides on
Creaming . .
Confirmation of the Mertens and Samuelsson
Effects. . . . .
Examination of the Role of K- Casein in Homogeni-
zation- Induced Destruction of Creaming . .
Examination of the Participation of Skim Milk
Membrane in Creaming
Effect on Creaming of Homogenized Skim Milk
Containing Washed Milk Fat Globules.
Effect on Creaming of Heated Skim Milk
Containing Washed Milk Fat Globules. . .
Effect on Creaming of Gravity- -Separated Skim
Milk Containing Washed Milk Fat Globules
Creaming in Mixtures of Gravity—Separated,
Homogenized, and Heated Skim Milks Con-
taining Washed Milk Fat Globules
Effect on Creaming of Raw Milk
Effect on Creaming of Cold Milk. . . . . .
Isolation of Skim Milk Membrane from Warm-
and Cold-Separated Skim Milks. . . . . .
Isolation of Skim Milk Membrane from Non-
Homogenized and Homogenized Skim Milks and
Model Systems. . .

vi

Page

76
76

76

78
Bl
8l
87
89

91
9‘3

94

99
102
l02
104
108
108
llO

_l——l.._.l
—J-—l—J
Cacao.)

ll6

ll8

Page

Effect of Replacing Skim Milk Membrane with
Milk Fat Globule Membrane. . . . . . . . . lZl

Creaming in Model Systems. . . . . . . . . . 122
Simulated Whey Model Systems . . . . . . . . l22
Simulated Skim Milk Model Systems. . . . l24

Simulated Milk Ultrafiltrate Model Systems . 124
Interaction of Milk Fat Globules with Skim Milk

Membrane . . . l27
Interaction of Immunoglobulin. M with Skim Milk

Membrane . . . 129
Interpretation of. the Effect of. Environmental

Factors on Creaming. . . 129
A Model Representative of Fat Globule. Cluster-

ing. . . . . . . . . . . . . . . . . . . . . . I33

CONCLUSIONS. . . . . . . . . . . . . . . . . . . . . I40
RECOMMENDATIONS. . . . . . . . . . . . . . . . . . . I41
APPENDICES . . . . . . . . . . . . . . . . . . . . . I42
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . I46

vii

Table

l0

ll

12

LIST OF TABLES

The recovery and purification of IgM at
selected steps in the isolation scheme.

Molecular weights of IgM heavy and light chains
as determined by sodium dodecyl sulfate poly-
acrylamide gel electrophoresis as a function of
gel concentration . . . . . . .

The effect of isolated IgG, void volume, and
IgM fractions on creaming of heated milk.

The effect of IgM, IgG, and IgA specific anti-
sera on creaming of milk. . . . . . . . . . .

The effect of 2—mercaptoethanol on creaming of
milk.

The effect of IgM on creaming of milk

The effect of temperature on the dispersion of
IgM, IgG, and the void volume fraction.

Quantitation of IgM involved in creaming of
milk supplemented with IgM.

Evaluation of gravity-separated skim milk to
support creaming following the addition of
washed milk fat globules. . . . . . .

Evaluation of gravity-separated skim milk from
IgM-supplemented raw milk to support creaming
following the addition of washed milk fat
globules. . . . . . . . . . . . . . . . . .

The effect of IgM isolated from warm-separated-
"total" - or gravity-separated - "inactive"
skim milk on creaming of heated milk. . .

The effect of IgM- derived Fab fragments on
creaming of milk. . . . . . . . . . .

viii

Page

48

6O

65

67

67
69

73

75

77

79

8O

86

Table Page

l3 Creaming in reconstituted milk containing
synthetic milk fat globules. . . . . . . . . . 88

l4 The effect of soluble milk fat globule mem-
brane proteins on creaming of milk . . . . . . 9O

l5 The effect of anti-milk fat globule membrane
and anti-navy bean trypsin inhibitor hyperim-
mune anti-sera on creaming of heated milk. . . 92

T6 The effect of simple sugars on creaming of
milk . . . . . . . . . . . . . . . . . . . . . 95

l7 The effect of enzymic treatment of milk fat
globules on creaming of milk . . . . . . . . . 100

18 The effect of milk fat globule membrane
gangliosides on creaming of milk . . . . . . . 103

l9 Confirmation of the Mertens and Samuelsson
effects - creaming in heated skim milk,
homogenized skim milk, or a mixture of heated
skim milk and homogenized skim milk containing
washed milk fat globules . . . . . . . . . . . 105

20 Creaming in homogenized and non-homogenized
acid and rennet wheys containing washed milk
fat globules . . . . . . . . . . . . . . . . . 107

21 The effect of skim milk membrane isolated by
ultracentrifugation or salt fractionation and
gel filtration on creaming of homogenized skim
milk containing washed milk fat globules . . . 109

22 The effect of skim milk membrane isolated by
ultracentrifugation on creaming of heated skim
milk containing washed milk fat globules . . . III

23 The effect of skim milk membrane isolated by
ultracentrifugation or salt fractionation and
gel filtration on creaming of gravity-separa-
ted skim milk containing washed milk fat
globules . . . . . . . . . . . . . . . . . . . 112

24 Creaming in mixtures of gravity-separated,

homogenized, and heated skim milks containing
washed milk fat globules. . . . . . . . . . . “4

ix

Table Page

25 The effect of skim milk membrane isolated by
ultracentrifugation on creaming of milk. . . . ll5

26 The effect of skim milk membrane isolated by
ultracentrifugation on creaming of cold milk . ll7

27 Creaming in simulated whey model systems . . . l23

28 Creaming in a simulated skim milk model
system . . . . . . . . . . . . . . . . . . . . 125

29 Creaming in a simulated milk Ultrafiltrate
model system . . . . . . . . . . . . . . . . . 126

30 The effect of IgM and skim milk membrane on
creaming of heated milk and heated milk fat
globule-washed milk. . . . . . . . . . . . . . 128

Al Chemicals used in this study and their
sources. . . . . . . . . . . . . . . . . . . . l42

A2 Equipment routinely used in this study . . . . 144

A3 Activities of glycosidases from Turbo cornutus
as supplied by Miles Laboratories, Inc.. . . . I45

Figure

IO

ll

LIST OF FIGURES

Schematic representation of the isolation pro-
cedure for obtaining IgM from skim milk.

Gel filtration chromatogram of a crude immuno-
globulin preparation

Gel filtration chromatogram of IgM .

Electrophoretic patterns characterizing the
isolation of IgM from milk

Sodium dodecyl sulfate polyacrylamide gel
electropherograms of molecular weight stan-
dard proteins and IgM.

Plot of molecular weight versus retardation
coefficient derived from sodium dodecyl
sulfate polyacrylamide gel electrophoresis
data . . . . . . . .

Two dimensional double-diffusion patterns of
pooled gel filtration fractions 44-55 (see
Figure 2) developed against (A) anti-IgM
anti-sera (u-chain specific), and (B)
anti-IgG anti-sera (y-chain specific).

Gel filtration chromatogram of alkylated,
2-mercaptoethanol-reduced, trypsin-treated
IgM. 0 O O 0 O O O O 0 O O 0 O O O O 0

Sodium dodecyl sulfate polyacrylamide gel
electropherograms of Fab fragments and IgM

Gel filtration chromatograms of (A) warm-
separated skim milk and (B) cold-separated
skim milk. . . .

A model representing an interpretation of

milk fat globule clustering as presented in
the contemporary literature.

xi

Page

25

49
52

55

58

61

63

82

84

ll9

I34

Figure Page

l2 A proposed model of milk fat globule clustering
based on the experiments and discourse of this
dissertation. . . . . . . . . . . . . . . . . . I36

13 A pictorial presentation of events occurring in
cold-agitated milk in the absence and presence
of supplemental skim milk . . . . . . . . . . . l37

INTRODUCTION

The rapid rising of fat globules with a subsequent
formation of a cream layer in normal cows' milk upon
quiescent cold storage represents one of the fundamental
physical properties of the milk system. This phenomenon
is referred to as "creaming". Despite the number of
investigations which have focused on this process, a Clear-
cut conception of the fundamental principles involved has
not been established.

In the days of creamline milk, the consumer equated a
deep cream layer with product quality. Early processors of
dairy products used creaming as a means of obtaining a
cream fraction from milk. With the introduction of effi-
cient mechanical milk separators and because creaming is
destroyed by heat treatment or homogenization - processes
which virtually all commercial milk undergo - creaming is
no longer of practical significance to the dairy industry.
Nevertheless, elucidating the mechanism of creaming remains
as an intriguing challenge.

Early studies demonstrated that fat globule clustering
wasa prerequisite for normal creaming. Immunoglobulin M
(IgM) has been shown to promote fat globule clustering.

Based on studies with IgM isolated from milk, the

immunoglobulin was designated as a cryoglobulin, i.e.,
undergoes cold-induced aggregation and precipitation. The
cold-induced aggregation and the concomitant non-specific
precipitation upon more than one fat globule, forming IgM
"bridges", is proposed to result in clustering. This theory
precludes an active role by other milk constituents and in
particular the milk fat globule or its membrane. Although
consistent with some of the observed phenomena, the theory
fails to account for phenomena such as the Mertens and
Samuelsson effects. In addition, the procedure employed in
assessing the apparent cold-induced aggregation of purified
IgM do not rule out a concentration and purity dependent
phenomenon. This property is characteristic of purified
immunoglobulins and in particular IgM and IgA. These
factors prompted us to re-investigate the mechanism of fat
globule clustering.

Research was directed toward determining the components
involved in fat globule clustering and examining whether
their role in creaming is passive or active. The nature of
the interactions and their heat and homogenization lability
was also examined. The ultimate goal of this study was to
develop a model representative of fat globule clustering

consistent with observed phenomena.

LITERATURE REVIEW

The Role of Fat Globule Clustering in Creaming

 

Babcock (l889a, l889b) was the first to call attention
to the clusters of fat globules which form shortly after
milk is drawn from the cow and cooled. He postulated that
the coagulation of fibrin, which he believed to be a normal
constituent of milk, entangled the fat globules and other
"solid and gelatinous“ matter in milk and weighed them down
enough to prevent them from rising. To minimize fibrin
clotting or to maximize cream separation he suggested that
milk be set at a cold temperature (ice water) immediately
after milking. This theory was not generally accepted and
was proven untenable when Hekma (l922) demonstrated that
fibrin is not a normal constituent of milk.

Babcock and Russell (l896a, 1896b) and Noll et__l.
(l903) observed that clustering is an important factor
influencing the consistency or viscosity of pasteurized milk
and cream. Whereas raw milk or cream contained fat globule
clusters, heated milk or cream failed to contain clusters.
They did not indicate a correlation between fat globule

clustering and creaming properties. However, their observa-

tions led to further investigations of fat globule clustering

and the formulation of numerous theories to explain the

phenomenon.

 

Henseval (1902) studied two types of milk, one showing a
rapid and one showing a slow rising of cream. Both types
had a normal composition but the milk which creamed rapidly
had numerous large fat globule clusters while the fat
globules in the slow creaming milk rarely showed any clus-
tering.

In a study of the efficiency of creaming in the deep-
setting and shallow-pan creaming methods, McKay and Larsen
(l906) found a more rapid and complete creaming using the
deep-setting method. This was attributed to the more
favorable conditions provided for aggregation of the fat
globules into "small bunches or masses.”

Hammer (l9l6) attributed the loss of cream layer forma-
tion in milk having undergone agitation or pasteurization
to the modification or size reduction of fat globule
clusters.

Van der Burg (l92l) called attention to the fact that
individual fat globules should rise according to Stokes'
law. The law which expresses the terminal velocity of small
spheres submerged in a viscous fluid had been successfully
applied to many systems. Rahn (l921), Van Dam and Sirks
(l922), and Troy and Sharp (T928) confirmed that fat glo-
bules in milk do rise according to Stokes' law. The rate
of rise of single globules was such that at least 50 hours
would be required for most of the fat globules to rise from

the bottom to the top of the creaming vessel. The rate of

rise of fat globule clusters agreed very well with that
calculated by Stokes' law, and was rapid enough to account
for the normal creaming time. These authors concluded that
the fat globules in raw milk do not rise singly but that
they aggregate into clusters which have considerably greater
velocities of rise. Their work conclusively established
that the clustering of fat globules plays the major role in
explaining the creaming phenomenon.

A technique introduced by Dahlberg and Marquardt (l929),
whereby thin films of milk are viewed with transmitted
light, added indisputable evidence to the proof of the
importance of fat globule cluster formation in gravity

creaming.

Components Involved in Fat Globule Clustering

 

Early investigators realized that the percentage of fat
in milk does not entirely determine its creaming ability
(Hammer, l9l6). The fact that milk with a high fat content
occasionally produced a thin cream layer and the variability
in creaming characteristics of milk from individual cows
made it quite evident that factors other than the percentage
of fat influence the creaming ability.

As discussed in the previous section, Babcock (l889a)
incorrectly attributed the differences in creaming charac—

teristics to a variable fibrin content.

Marcas (I903, I904) attributed the poor performance of
slow creaming milk to its elevated ash, phosphoric acid, and
lime concentrations.

VanEMm_;t_;l. (I923) found that the addition of blood
serum to milk which creamed poorly improved its creaming
properties. If blood serum had been heated to 65 C or
higher,little or no improvement was observed. The authors
suggested that the clustering of fat globules in milk may be
due to an agglutinin common to both milk and blood. Since
milk is elaborated by the mammary gland from materials
brought to it by the blood, these authors considered it con-
ceivable that the blood serum from cows whose milk creamed
well, when added to poor creaming milk, would influence the
creaming more favorably than the blood serum from cows whose
milk creamed poorly. Experiments failed to establish such a
relationship. Brouwer (I924) later demonstrated that blood
serum from steers was as beneficial in promoting creaming as
that from cows.

In a continuation of these experiments, Hekma and Sirks
(l923) examined the influence of various serum fractions on
creaming. The substance responsible for the improvement in
creaming was precipitated with sodium chloride or ammonium
sulfate almost completely along with the globulin fraction.
When the globulin fraction was dialyzed and added to milk,
the creaming properties were almost identical to those of a

control sample containing an equivalent volume of added

 

serum.

Brouwer (I924) fractionated the globulin fraction into
euglobulin and pseudoglobulin fractions and demonstrated
that euglobulin was the active component. This was suppor—
ted by showing that blood serum of new born calves, which
is devoid of euglobulin, failed to improve creaming proper—
ties.

Palmer _t _1, (I926) and Troy and Sharp (I928) demon-
strated that the agglutinin was present in the whey fraction
of milk as expected.

Orla-Jensen (I929) attributed the small amounts of
colostrum required to restore creaming to heated milk to its
high globulin concentration.

Rowland (I937) provided further support for the role of
the globulin fraction in whey by showing that the reduction
in creaming power is proportional to the percentage of total
albumin and globulin denatured in heated milk.

Sharp and Krukovsky (I939) found that the agglutinin
was concentrated in cream separated at low temperatures and
was relatively absent in the corresponding skim milk. The
opposite was true for milk separated at high temperatures.
By first separating at a low temperature and then further
separating the cream to a higher fat content at a higher
temperature at which the fat is liquid, a cream plasma very
active in agglutinating power was obtained. They concluded

that the agglutinating substance normally present in milk is

adsorbed on the surface of solid fat globules but not on
liquid fat globules.

Dunkley and Sommer (I944) found that the agglutinin
nearly quantitatively precipitated from agglutinin-rich
wheys when allowed to stand undisturbed for I2 hours at 5 C.
Agglutinin-poor wheys gave very little or no precipitate. A
euglobulin fraction prepared from the proteins in the pre—
cipitate produced practically the same creaming behavior as
an equivalent amount of the precipitate indicating that the
agglutinating power resides in the euglobulin fraction.

Samuelsson _t al. (l954) also observed that the agglu—
tinin responsible for normal creaming in milk precipatated
at 2 C from rennet whey which had been heated to not greater
than 60 C. The agglutinin formed opalescent solutions in
warm whey or water and could be precipitated by adding
sodium chloride or gum arabic. The agglutinin was shown to
consist of two components, one inactivated by homogenization
and the other by heating. In systems containing cream mixed
with water, whey, or separated milk creaming would occur if
one portion of the available agglutinin had been inactivated
by homogenization and the other by heating, but if all the'
agglutinin had undergone either of these treatments, no
creaming resulted. The activity of agglutinin was favored
by factors contributing to its gradual precipitation, e.g.,
dilution, low—temperature aging in the range of 5 to I0 C,

and mild heat treatment. Similar observations were reported

 

by Kenyon and Jenness (I958).

After concentrating the active clustering agent from
milk in a cream plasma according to the procedure of Sharp
and Krukovsky (I939), Gammack and Gupta (I970) subjected
the aqueous phase to ultracentrifugation. They recovered
85% of the clustering activity in a small opalescent layer
which sedimented above the casein pellet. A precipitate
which formed when the opalescent layer was diluted with milk
Ultrafiltrate and held at 4 C contained most of the clus-
tering activity. The precipitate could be redispersed on
warming. Although the euglobulin fraction from colostrum
also exhibited the cryoprecipitation behavior, precipitate
from the opalescent layer was twenty times more active in
terms of its protein content in promoting clustering and
caused a more rapid formation of cream line than did euglo-
bulin. When the precipitate was fractionated by gel fil-
tration, the clustering activity was localized in a high
particle-weight fraction, i.e., greater than IxI06, which
contained lip0protein particles. The isolated lip0protein
particles themselves showed no clustering activity.
Immunoglobulin G (IgG) isolated from colostral euglobulin
had no clustering activity with or without added lip0protein
particles. A fraction containing immunoglobulin M (IgM) and
A (IgA) isolated from colostral euglobulin showed some
activity with a slow formation of cream line. On addition

of lipoprotein particles,clustering activity was markedly

lO

enhanced and cream line formation was rapid. These authors
concluded that lip0protein particles which exist in the
aqueous phase augment the clustering activity of immune
proteins such as those in euglobulin properties. In a set

of related experiments, Franzen (l97l) was not able to obtain
similar results. This may have been due to different cen-
trifugation conditions or the use of heated skim milk.

The observations of Gammack and Gupta support the
earlier work of Hansson (I949) who noted that creaming in raw
or low-temperature treated milk was enhanced by the addition
of phospholipids isolated from brain matter. Cream rising
was not improved in high-temperature treated milk.

Payens and Both (I970) and Franzen (l97l) have also

demonstrated that the IgM fraction is the euglobulin compo-
nent active in restoring creaming to heated milk.

Stadhouders and Hup (I970) and Bottazzi 33 _1. (I972)
have suggested that there are 3 different types of aggluti-
nins present in milk, binding either fat globules together,
bacteria together, or bacteria to fat globules. Stadhouders
and Hup demonstrated that the fraction which separates from
euglobulin at low temperatures contains the immune proteins
which agglutinate fat globules and those which attach the
bacteria to fat globules.

Bottazzi and Premi (I977) contend that fat globules

that remain in the skim phase during the agglutination pro-

cess have a higher level of 5'-nucleotidase than the fat

ll

globules which rise. They suggested fat globules that agglu—
tinate most quickly are those with a low content of phosphor
lipids.

In a study of factors associated with production of
milk with a low creaming capacity, Bertoni gt _1. (I979)
found milk with a high creaming capacity contained high
levels of fat but low levels of phospholipids and cholesterol.
They interpreted these results as indicating that fat globule
membranes from milk with a high creaming capacity were less
compact.

Bottazzi and Zacconi (I980) have recently isolated from
milk a component active in aggregation of fat globules and
bacteria using a new procedure. After concentrating the
active clustering agent from milk in a cream plasma according
to the procedure of Sharp and Krukovsky (I939), the aqueous
phase was frozen. After thawing, a cryofloculate recovered
by centrifugation was fractionated by gel filtration and
ultracentrifugation. A fraction which greatly increased the
degree and rate of cream rising was obtained. Virtually no
cryoflocculate was observed in milk with a low creaming

capacity.

Physical and Chemical Factors Influencing Creaming

 

Low Temperature

 

A low temperature has generally been recognized as a

prerequisite of normal creaming. As discussed previously,

l2

Babcock (l889a) incorrectly attributed this to the prevention
of fibrin coagulation. Later investigators (Rahn, I92I;
Vanikm1and Sirks, l922; Troy and Sharp, I928) recognized the
role of decreased temperatures in promoting cluster forma-
tion. Sharp and Krukovsky (I939) made the significant obser-
vation that agglutinin material was preferentially associated
with fat globules in the cold and the skim phase in warm
milk. Using labeled euglobulin, Payens gt _1. (I965) conclu-
sively demonstrated that the amount adsorbed on fat globules
is strongly temperature dependent. Reduced temperatures
increased clustering of the fat globules and enhanced;
adsorption of euglobulin.

Dunkley and Sommer (I944) and Samuelsson et_al. (I954)
noted that factors which favor precipitation of the active
component from whey also favor fat globule clustering. Rhee
(I969) and Franzen (l97l) associated fat globule clustering
with temperatures low enough to promote euglobulin or IgM
aggregation,respectively. Payens and Both (l970) suggested

that the same functional groups are responsible for IgM

cryoaggregation and fat globule cold agglutination.

High Temperature

 

Many of the early studies were initiated to examine the
reduction in creaming capacity due to excessive heat (high
temperature pasteurization). The reduced creaming power of
heated milk was initially believed to be due to the precipi-

tation of denatured lactalbumin onto fat globules resulting

13

in their failure to rise (Hunziker, l92l). Later investi-
gators (Rahn, l92l; Van Dam and Sirks, I922; Troy and Sharp,
I928) established experimentally that the reduction in cream-
ing properties of milk caused by the use of high pasteurizing
temperatures results from interference with subsequent clus-
ter formation. Hekma and Sirks (l923) attributed the lack
of clustering to the heat lability of the agglutinin respon-
sible for creaming. Palmer and Anderson (I926) demonstrated
that pasteurization primarily affected the milk plasma.
Rowland (I937) supported this contention by showing that the
reduction in creaming power was proportional to the percen-
tage of the total albumin and globulin denatured. Based on
electrophoretic studies with fat globules in samples of
heated whole milk, heated skim milk plus raw cream, and raw
skim milk plus heated cream, Dahle and Jack (I937) conclusi-
vely established that the electrokinetic potential of milk
fat globules is not a major factor in the reduction of
creaming by heat treatment. These studies were supported

by experiments demonstrating the restoration of creaming to
heated milk upon addition of euglobulin (Dunkley and Sommer,
I944) or IgM (Payens and Both, I970; Franzen, l97l).

Whereas high-temperature treatment eliminates creaming
capacity, Dunkley and Sommer (I944) demonstrated that it was
necessary to warm (50 C) the fraction (skim or cream) con-
taining active agglutinin for normal creaming to be observed

in recombined samples. It was necessary to warm only that

 

l4

portion which contained the agglutinin. Normal creaming was
only restored in whole milk samples by warming prior to

placement at reduced temperatures.

Homogenization

 

Hammer (l9l6) attributed the loss of creaming capacity
in homogenized whole milk to the reduced size of fat globules
which have less tendency to rise than the original globules.
Doubt was shed on this theory when Mertens (I932) observed
that milk recombined from normal cream and homogenized skim
milk did not cream. Experiments conducted by Dunkley and
Sommer (I944) confirmed these results. Furthermore, it was
shown that milk recombined from unclumped homogenized cream
and normal skim milk formed a cream layer. They attributed
the loss of creaming capacity in homogenized milk to the
denaturation of agglutinin. Samuelsson _£.il- (I954) demon-
strated that the agglutinin consisted of a heat- and
homogenization-labile component. Payens (I964) was unable
to detect any differences in the physical properties and
clustering ability of euglobulin isolated from colostrum
prior to and following homogenization. In a subsequent
study, Koops et_al. (I966) observed normal creaming when
washed cream was added to a homogenized model system contain-
ing milk dialyzate and euglobulin. However, when micellar

casein or K-casein was added to the system, creaming was

eliminated despite the observation that labeled euglobulin

l5

was adsorbed on the fat globule surface. The adsorbed
euglobulin was accompanied by small amounts of casein,
particularly K-casein. The authors suggested that a euglo—
bulin-K-casein complex was formed during homogenization
which was capable of adsorbing to the fat surface but
unable to effect clustering, or that homogenization induced
the adsorption of K-casein on the fat surface, screening
adsorption sites for euglobulin. Walstra (I980) homogenized
rennet and acid whey (after pH adjustment to 6.6) and found
creaming capacity was fully destroyed in both samples. The
formation of a euglobulin-K-casein complex was therefore

doubted.

E

In confirmation of previous work, Dunkley and Sommer
(l944) found alkali added to milk increased the depth of the
cream layer and resulted in a more rapid and complete cream-

ing. Acid was found to have an opposite effect.

Added Salts

 

Dunkley and Sommer (I944) found the addition of sodium
citrate and disodium phosphate to milk to have little effect
on cream volume. Increasing concentrations of sodium
chloride, calcium chloride, aluminum chloride, and ferric
chloride caused marked reductions in the depth of the cream
layer. The authors therefore suggested that the charge on

the fat globule is not an important factor in determining

16

the creaming properties of milk.

Creaming in Milks Other than Cows'

 

The slow rate of creaming in buffaloes' milk has been
attributed to the absence of cluster formation (Fahmi,
I951). Fahmi e£__l. (l956a, l956b) supported this notion
and reported that goats' and sheeps' milk are also devoid
of fat globule clusters at room (22 C) or refrigerator
(I0 C) temperatures. Abo-Elnaga _t 31. (I966) reported
clusters in buffaloes' milk held at 4 C for 3 hours but
that they rarely contained more than 4 fat globules. They
attributed the poor creaming of buffaloes' milk to low
levels of agglutinin.

After a study of the factors influencing the creaming
ability of carabaos' milk, Gonzales-Janolino (I968) conclu-
ded that carabaos' milk lacks the homogenization-labile
component and while the heat-sensitive component is present,
it is not in sufficient quantities for normal creaming. The
poor creaming carabaos' milk was found to cream to an extent
at least as exhaustive as that in cows' milk upon addition
of an agglutinin concentrate from cows' milk.

Jenness and Parkash (l97l) attributed the poor creaming
of goats' milk at low temperatures to the lack of aggluti-
nating euglobulins. Milks reconstituted from goat's cream

and cows' skim milk creamed readily whereas those made

by combining cows' cream and goats' skim creamed poorly.

 

l7

The goats' milk fat globules failed to cream as exhaustively
as those of cows' milk.

Whittlestone (I953) showed that creaming occurs in
mixtures of sows' cream and cows' skim milk held at 37 C for
2-I6 hours while sows' milk failed to exhibit cream forma-
tion under these conditions. Although the experiment was
not performed at an optimal temperature, the author sugges-
ted that the agglutinin of cows' milk will cause the clus—
tering of sows' milk fat globules. These experiments

indicate sows' milk lacks a fat globule agglutinin.

Mechanisms of Fat Globule Clustering

 

The recognition of the significance of fat globule
clustering in promoting cream layer formation led to numer-
ous investigations and theories to explain the phenomenon.
Early theories, which were critically reviewed by Dunkley
and Sommer (l944), were based on (a) gravitation of fat
globules, (b) electrokinetic potential of fat globules,

(c) interfacial tension, (d) stickiness and state of hydra-
tion of the adsorbed membrane, and (e) fat clustering
considered as an agglutination process. Based on experi—
mentation designed to test these theories Dunkley and Sommer
concluded that (a) the variable creaming properties of
normal milk cannot be explained on the basis of differences
in the rates of rise of individual fat globules, (b) the

salts normally present in milk are sufficient to reduce the

l8

surface potential on the fat globules and thus eliminate
the charge variability of the fat globules, (c) the inter-
facial tension at the fat-serum interface does not promote
creaming, and (d) evidence concerning the importance of
hydration of the membrane on the fat globules was not suf-
ficient to justify drawing a definite conclusion regarding
the significance of this factor. Fat globule clustering was
noted to share many similarities with bacterial aggluti-
nation. Both processes involve the aggregation of particles,
require the presence of globulins, are prevented by heat
denaturation, and require optimum salt concentrations. They
used Marrack's (I938) theory of bacterial agglutination as a
model and stated that if the mechanisms were similar, clus-
tering would be promoted by (a) a partial dehydration of the
adsorbed membrane on the fat globules due to specific polar
adsorption of euglobulins, (b) aggregation of fat globules
resulting from the adsorption of a single euglobulin mole-
cule by two fat globules, and (c) maintenance of the surface
potential of the fat globules below the critical level by the
presence of salts. It was concluded that the clustering of
fat globules in milk takes place by the same mechanism as
that involved in the agglutination of bacteria.

Brunner (I974) has stated that it is unfortunate that
this comparison has been made since it implies the operation
of an antibody-antigen interaction. No evidence exists to

suggest that the fat globule membrane contains antigenic

l9

components to the euglobulin fraction. The results of Stad-
houders and Hup (I970) indicating a lack of specificity in
the euglobulin—fat globule complex but a specificity in the
antibody-bacteria complex supports this suggestion. As the
mechanism of bacterial agglutination was a matter of con-
troversy in I944 and data supporting fat clustering as an
agglutination process was only based on comparison of
limited aspects of the two phenomena, the use of the term
agglutination for fat globule clustering can only be justi-
fied on the basis of convenience. Dunkley and Sommer's
actual theory embracing a physical, as opposed to a specific,
interaction between the euglobulin components and the fat
globule surface has been generally accepted.

Jenness and Patton (I959) suggested that the adsorbed
euglobulins serve to reduce the fat/plasma interfacial
tension, thus permitting the globules to approach one another
to form clusters. The fact that euglobulin-rich skim milk
foams copiously was offered as evidence in support of the
interfacial activity of euglobulin. Actually, cold sepa-
rated skim milk has a lower surface tension than euglobulin-
rich warm separated skim milk (Brunner, I974). Furthermore,
Jackson and Pallansch (1961) demonstrated that euglobu-
lin is not as interfacially active as other milk proteins
such as o-lactalbumin, and B-Iactoglobulin, or the native
interfacial milk fat globule membrane protein. It therefore

seems improbable that the clustering of fat globules can be

20

explained solely on the basis of an increase in the inter-
facial activity resulting from the adsorption of euglobulin
at the interfacial surface.

Payens (I964) suggested that the adsorbed euglobulin
would cause agglutination by reinforcing the London-Van der
Waals attraction between the fat globules. In view of the
small amount of euglobulin experimentally determined to be
adsorbed to the fat globules, this suggestion was retracted
(Payens _§._l., I965). Rather it was suggested that aggluti-
nation is brought about by the formation of euglobulin bridges
between the fat globules, essentially in accordance with
the model proposed previously by Dunkley and Sommer.

Kenyon _t ”1. (I966) attributed the association of
euglobulins with the fat globule surface to ionic interac-
tions promoted by euglobulins having an isoelectric point
relatively more positive than the fat globule surface. It
was suggested that at low temperatures a more rigid system
may develop forming lipoproteins held by weak ionic bonds
with the basic proteins. Larger molecular weight proteins,
such as euglobulins, would have an extended sphere of attrac-
tion thus more readily forming bridges with the ionic
structure of phospholipids on adjacent fat globules. The
Brucella Ring Test was postulated to involve immunoglobulins
involved in fat globule clustering and having specific

antibody activity (for Brucella cells). Adsorption to fat

globules was proposed to occur in such a way that their

2l

combining sites remain available for reaction and hence still
capable of agglutinating bacteria.

The most recent and the generally accepted theory
(Brunner, I974; Mulder and Walstra, I974) attributes agglu-
tination to the precipitation of cryoglobulins (IgM) onto
the fat globules, leading to clustering and subsequently to
rapid creaming of the large clusters (Payens, I968; Payens
and Both, I970). Since several observed phenomena are diffi-
cult to explain by the simple mechanism, Brunner (I974) and
Mulder and Walstra (I974) have suggested that the clustering

reaction is more complicated.

Facets of Fat Globule Clustering Warranting

 

Further Investigation

 

Before a theory of fat globule clustering consistent
with observed phenomena can be proposed, numerous questions
regarding the nature of the components involved must be
answered or further examined. Does all the IgM in milk
undergo cryoprecipitation? Why does it associate only with
the fat globules - is the reaction specific or non-specific?
00 fat globules only play a passive role as has generally
been assumed? What is the nature of the homogenization-
labile component? No mechanism to date has prescribed a
role to this component - how is it involved in fat globule
clustering? How do the lip0protein particles of Gammack and

Gupta augment creaming?

22

The conditions required for fat globule clustering
also warrant further investigation. Why is a low tempera-
ture prerequisite for creaming? Is only IgM effected? Do
salts play an indirect or a direct role in the process?

What component(s) involved in the process is effected as the
pH is varied?

It is of interest that of the milks examined only the
fat globules of cows' milk display extensive clustering.

The other milks appear to contain the components required
for creaming - euglobulins (IgM) and fat globules. What is
unique about the IgM or fat globules in cows' milk? What is
the biological significance with respect to immunoglobulin

synthesis or milk secretion?

 

EXPERIMENTAL

Materials and quipment

 

The milk used in this study was obtained from the
Michigan State University Holstein dairy herd. Milk was
collected in three or five gallon stainless steel cans from
the milking parlor and separated as soon as possible at
40-45 C.

Rabbit anti-sera against bovine IgM, IgG, and IgA
were purchased from Miles Laboratories, Inc. Rabbit anti—
sera against bovine xanthine oxidase, bovine milk fat
globule membrane, and navy bean (E. vulgaris L.) trypsin
inhibitor were provided by Dr. J.R. Brunner.

Trypsin, type III, neuraminidase, type V, and rennet,
type I, were purchased from Sigma Chemical Co. Pronase,
grade B, was obtained from Calbiochem-Behring Corp. and
mixed glycosidases from Miles Laboratories, Inc.

Bio-Gel A-0.5n1 and A-5n1were obtained from Bio-Rad

Laboratories; Sepharose 4B and Sephacryl S-200 from Pharmacia

Fine Chemicals; and agarose (-mr less than 0.2) from Miles
Laboratories, Inc.

Chemicals used in this study along with their sources
are listed in Table Al of the Appendix. All chemicals were

reagent grade unless otherwise indicated. Equipment used

23

 

24

regularly during the course of this study is listed in
Table A2 of the Appendix. Instrumentation specific for a
certain experiment will be referred to in the appropriate

section.

Preparative Procedures

 

Immunoglobulin M

 

Immunoglobulin M (IgM) was prepared by combining the

 

ammonium sulfate precipitation method described by Smith
(l946a, l946b) and gel filtration on Bio-Gel A-51nand
A-0.51ncolumns connected in tandem. Figure I outlines the
isolation procedure of IgM from fresh skim milk. Whole milk
was separated in a laboratory disk separator at 40-45 C.
The resulting skim milk was cooled to room temperature and
adjusted to pH 4.6 with IMIiCI and allowed to stand I h.
The precipitated casein was removed by filtration through 3
layers of cheese cloth. The acid whey was adjusted to pH
6.5 with l M NaOH and solid ammonium sulfate was added slowly
to 50% saturation (3I3 g/l) to salt-out the crude lactoglo-
bulin fraction. After standing overnight, most of the
supernatant was siphoned off, and the precipitate was
collected by centrifugation at l6,000)<g for 20 min at 20 C.
The precipitate was solubilized in a 0.3llNaCl solu-
tion containing 0.02% (w/v) NaN3 at approximately a 3% (w/v)
protein concentration. The pH was adjusted to 4.6 with l M
HCI and ammonium sulfate added to 25% saturation (I44 g/l).

After centrifuging at l6,000><g for 20 min at 20 C, the

25

Acre om com o xooo o

va :m new chmu Am pcwcanoo
.cwmm &o 4 cu comm Ad I: : nw< mmopamaimmompoca
o. o op In .cwmmmu Fmsvwmmmv
Hz<e<zmma2m mH<H~dHomma
. a

 

Aces om com a xooo.owv mmzwrcpcmu

0mm 412V ee<
o. e on Ia
:ovpmspcmocoo xm op Fonz z m.o cw m>_ommvo

.epam emm on

 

Acwszapm muscuv Acw_:no_m muscuv

 

 

 

0 mm ou pmznn<

ez<e<zmma2m me<eHaHumma
. .
Acws om com a xooo.owv mozwecpcmu
.eoem a0m op 0mm 412v ee<
m. o on In
Azmszv Aswwmmov
ez<e<zmma=m me<eHaHomma
. - 15
ea
o.e on In

XJHZ szm

26

 

.JFVE wam
gone zmH mcwcwmpno toe mczcmuocq coepm_omw mcp to :owpmpcmmmcawc owpmsmzum ._ mczmwd

zoaeu<md eIaHmz m<b=um402 gob
zoaeu<md zmH
oneu<md m2:n©> DHo>

:owpmeprwm Pow
coo_ea
Loeezn cE:_oo c? wNwrwn:_om

 

Hz<H<Mmma3m mH<HHfiHummm

Acne om toe m xooo.emv oozetcoeou
.eoem sow op omNAeIzv ee<

mm><4 szHomaoaHA pz<b<zm3m

_ 5

Ac; m coe m xooo.oo_v oaseaeoeou
N._ eo spameoo a on cmez ee<
wawzn :Espoo cw w>_ommwo

 

Amcw—zao_moczesw

mcwvzpuxm Amcw_:no_mocssew
mcvmpoga xmcz _m:vvmmmv evacuv
Hz<p<zmmasm mH<HHmHumma
— _

 

27

supernatant was filtered through a single layer of cheese
cloth. A crude immunoglobulin fraction was precipitated
from this supernatant by adjusting the pH to 6.0 with I M
NaOH and adding ammonium sulfate to 40% saturation (99 g/l).

The crude immunoglobulin fraction collected by centri-
fugation at l6,000x g for 20 min at 20 C was dissolved in
column buffer- 0.0I M Tris-HCI, pH 8.0, containing 0.30 M
NaCl and 0.02% (w/v) NaN3- at approximately a 3% (w/v) pro-
tein concentration. In order to remove lipoproteins and
residual lipid material, NaBr was added to a density of I.2
(257 g/I) and the sample centrifuged at l00,000x g for 3 h
at 20 C. The subnatant was removed with a syringe and
needle to which ammonium sulfate was added to 50% saturation.
The crude immunoglobulin fraction was stored as such at 4 C
until required.

Prior to gel filtration the precipitate was collected
by centrifuging at 27,000x g for 30 min at 20 C. The preci-
pitate, representing approximately 2 l of whey, was dissolved
in 8 ml of column buffer and filtered through a 511 Milli-
pore filter. Gel filtration was performed in the ascending
mode at room temperature using columns of Bio-Gel A—5 m
(2.6 cm x 65 cm) and A—0.5 m (2.6 cm x 35 cm) connected in
tandem. Pre-swollen agarose gels from Bio-Rad Laboratories
were prepared and packed according to the manufacturer's
recommended procedures. Columns and 3-way and 4-way valves

were purchased from Pharmacia Fine Chemicals. The sample

 

28

applied to the A-5n1column at a rate of 0.20nfl/nflr1was fol-
lowed by 5 ml of a I0% sucrose solution prepared in column
buffer and applied at the same rate. Elution was achieved at a rate
of 0.35 ml/min. The column eluent was monitored at 254 nm
with an ISCO Ultraviolet Analyzer, model Uv-2. An ISCO,

model ll00, fraction collector was used for collecting ll

min fractions. A Beckman Instruments DK—2A spectrophotometer
was employed to read the absorbance of fractions at 280 nm.
When required, IgM concentration was determined from absor-
bance at 280 nm and a specific absorption coefficient

(E280 nm) of l2.6 (Mukkur and Froese, l97l).

Immunoglobulin M Fab Fragment

Immunoglobulin M Fab fragments were isolated according
to a procedure adapted from Beale and Buttress (I972).
Similar procedures have been shown to produce active Fab
fragments (Stone and Metzger, I968). A one-tenth volume of
2.5 M Tris-HCI, pH 8.0, buffer was added to IgM in column
buffer at a concentration of 0.80 mg/ml. After flushing
with nitrogen, dithiothreitol was added to a concentration
of 5 mM. This was allowed to react for l h at room tempera-
ture and iodoacetamide, recrystallized 4x from distilled
water, was then added to a concentration of 35 mM. After
reacting at room temperature for IS min, the solution was
dialyzed against several changes of column buffer for l8 h
at 4 C. Iodoacetamide and all solutions containing iodoace-

tamide were maintained in the dark. The solution was

 

29

concentrated approximately 4-fold with a Millipore Corpora-
tion Immersible CX unit (nominal molecular weight cut-off
l0,000 daltons).

After concentration and addition of CaClZ to a con-
centration of 0.0l M, the alkylated IgM subunits were
digested with trypsin (protein:enzyme weight ratio, 50:I)
for 3 h at 30 C. A 3-fold weight excess of soybean trypsin
inhibitor (I mg inhibits l.5 mg trypsin) was added to stop
digestion.

The Fab fragments were purified by gel filtration
chromatography performed in the descending mode at room
temperature using a column of Sephacryl S-200 (2.6 mnx
90 cm). Pre-swollen gel from Pharmacia Fine Chemicals was
prepared and packed according to the manufacturer's recommended
.procedures. The column buffer was the same as used for
IgM isolation - 0.0l M Tris-HCI, pH 8.0, containing 0.30 M
NaCl and 0.02% (w/v) NaN3. Before use, the column was
calibrated (log MW vs Kd) using Blue Dextran, vitamin B12,
bovine IgG, bovine serum albumin, ovalbumin, and sperm
whole myoglobin. A IO ml sample containing I0% sucrose and
representing approximately 30 mg of IgM, was applied to the
chromatographic bed under the eluent. A flow rate of 0.45
ml/min was employed. The column eluent was monitored as
for IgM isolation. A specific absorption coefficient

(E280 nm) of l2.6 was employed for the Fab fragment.

   

3O

Soluble and Insoluble Protein Fractjons of the Milk Fat

Globule Membrane

 

The procedure described by Herald and Brunner (I957)
was used to isolate soluble and insoluble protein fractions
from the milk fat globule membrane. This procedure was
chosen since it yields a soluble glycoprotein fraction more
representative of the total membrane fraction than other
procedures commonly employed (Yamauchi 33 gl., I978).
Approximately 8 gallons of fresh whole milk was separated
at 40-45 C. The cream was collected and washed by adding
3 volumes of distilled water at 40-45 C followed by gentle
mixing and reseparation. The cream was again collected
and distilled water was added to bring the volume back to
the original volume. This washing procedure was repeated 3
more times to ensure the complete removal of casein and whey
components from the cream.

Following storage overnight at 4 C the washed cream was
churned at room temperature. It required 40-60 min for the
emulsion to break, after which churning was continued until
the butter granules clumped to form large balls. The butter-
milk was separated from the butter granules by straining
through cheese cloth and an equal volume of saturated
ammonium sulfate solution was added to the buttermilk to
give a 50% saturated solution. This solution was allowed to
sit overnight at 4 C in a graduated cylinder during which

the salted-out membrane material formed a floating layer.

 

3l

After removing the subnatant, the floating membrane mate-
rial was centrifuged at 25,000x g for 60 min to concentrate
the membrane material into a packed,floating layer. The
resulting concentrated total membrane preparation was
delipidated as follows. Five ml of cold ethanol: diethyl
ether (35:65, v/v) per gram of concentrated membrane
material was added and the mixture stirred at 4 C for 60
min, followed by decantation and addition of an equal volume
of fresh solvent. After stirring for an additional 60 min
the membrane material was collected by centrifugation.

Five ml of cold diethyl ether was added per gram of original
membrane material. This mixture was stirred 60 min at 4 C.
The diethyl ether was decanted and replaced with an equal
volume of fresh diethyl ether. After stirring for 60 min
the membrane material was collected by centrifugation.
Fresh diethyl ether was added and the mixture stirred for
60 min at room temperature. The solvent was decanted and
an equal volume of fresh diethyl ether added. After stir-
ring for 60 min at room temperature the membrane material
was collected by centrifugation. Residual solvent was
removed under vacuum. The fat globule membrane proteins
were then fractionated into a soluble and an insoluble
fraction by extracting with 0.02 M NaCl and centrifuging
the mixture for 30 min at 25,000><g. This procedure was
repeated 2 times. The soluble and insoluble fractions were

dialyzed against distilled water for 48 h and Iyophilized.

32

Skim Milk Membrane

 

Preparation of skim milk membrane via preparative

ultracentrifugation. The procedure of Plantz et al. (I973)

 

was employed. Fresh whole milk warmed to 37 C was centri-
fuged at 5,000x.g for 20 min at 30 C. The skim milk was
removed with a syringe and needle inserted through the
compacted cream layer. The membrane material in skim milk
was obtained by centrifugation at 90,000x g for l00 min at

20 C. After removing most of the whey, the fluff material

 

located on the casein micelle pellet and tube bottom were
suspended by gently shaking the tube and collected with a
disposable pipette. Generally, 40 ml of skim milk membrane
(SMM) solution was obtained from 500 ml of skim milk. If
it was desired to obtain a SMM fraction with a lower whey
protein content, 3 volumes of simulated milk Ultrafiltrate
(Jenness and Koops, I962) was added and the mixture recen-
trifuged. Generally, l5 ml of SMM solution was obtained
from I60 ml of solution.

Preparation of skim milk membrane via salt fractiona-

 

tion and gel filtration. A procedure adapted from Kitchen

 

(I974) was employed. To skim milk, prepared as for prepara-
tion of SMM via ultracentrifugation, rennet was added at a
rate of 0.8 g/l. After 30 min at 30 C, the curd was cut

and allowed to sit an additional 30 min. The whey was then
removed by filtration through cheese cloth. The whey was

clarified by centrifugation at I6,000><g for 20 min at 20 C.

 

33

Ammonium sulfate was then added to 50% saturation. After
sitting overnight, most of the clear supernatant was siphoned
off and the precipitate collected by centrifugation at
l6,000)<g for 20 min at 20 C. The precipitate was solu-
bilized in 0.5 M NaCI and dialysed against several changes

of 0.5 M NaCl for 48 h at 4 C. After warming to 55 C for

30 min the solution was centrifuged at l5,000><g for 20 min
at 20 C.

The SMM was then purified by gel filtration chromato-
graphy performed in the descending mode at room temperature
using a column of Sepharose 4B (2.6 cm x 35 cm). Pre-swollen
agarose gel obtained from Pharmacia Fine Chemicals was pre-
pared and packed according to manufacturer's recommended
procedures. The column was equilibrated in 0.5 M NaCI.

Ten ml of sample containing l0% sucrose and representing

l50 ml of skim milk was applied to the chromatographic bed
under the eluent. A flow rate of 0.35 mI/min was employed.
The SMM was obtained in fractions which eluted at the void

volume. Column eluent was monitored as for IgM isolation.

Milk Fat Globule Membrane Gangliosides

 

A ganglioside-containing fraction was isolated from
milk fat globule membrane according to the procedure of
Keenan (I974). The gangliosides were not purified using
column or thin-layer chromatography. In addition to gangli-

osides, the aqueous fraction would contain neutral, mostly

34

fucose-containing, glycolipids, sulfatides, and some non-
lipid contaminants. However, it should not contain sialo-
glycoproteins (Brunngraber e£_gl., I976).

Milk fat globule membrane was prepared from washed
milk fat globules as described in the section on isolation
of soluble and insoluble milk fat globule membrane proteins.
The membrane material was collected by centrifugation at
90,000)<g for I00 min at 20 C. Lipids were extracted by
stirring the membrane material with 20 volumes of chloro-
formzmethanol (2:I, v/v) overnight at room temperature.
Insoluble residues were recovered and stirred with ID volumes
(based on original sample volume) of chloroformzmethanol
(l:l, v/v) for 6 h at room temperature. Combined extracts
were evaporated to near dryness with a rotary evaporator
and redissolved in chloroform:methanol (2 l, v/v). Ganglio-
sides were recovered by washing with 0.2 volumes of 0.88%
KCI and then once with theoretical upper phase without KCI
(chloroform:methanol:water::3:48:47, v/v/v). The combined
upper phases were then dialysed against several changes of

distilled water for 48 h at 4 C.

Casein Fractions

 

Pregaration of whole casein. Whole casein was pre—

 

pared from skim milk separated from whole milk at 40-45 C.
After cooling to room temperature the pH was slowly adjusted
to 4.6 with l M HCI while stirring. After sitting for I h,

casein was collected by filtration through 3 layers of cheese

 

35

cloth. The casein was washed 2 times with a volume of dis-
tilled water equal to that of the original milk. It was then

solubilized in an equal volume of distilled water by the slow

addition of l M NaOH to pH 7.0 and again precipitated and
washed 2 times with distilled water. After solubilization

and dialvsis against several changes of distilled water for

48 h at 4 C the casein was lyophilized.

Preparation of K-casein. K-Casein was prepared accor-

 

ding to the procedure of Zittle and Custer (I963). Approxi-
mately I50 9 of acid precipitated whole casein, prepared as
for whole casein isolation, was dissolved in 500 ml of

6.6 M urea. This solution was acidified with 100 ml of 7 N
H2804. After acidification, l I of distilled water was
added. After standing for 2 h the precipitate, containing
mainly as- and B-casein, was removed by centrifugation at
l6,000><g;for 20 min at 20 C. The K-casein was then pre-
cipitated by the addition of I32 9 ammonium sulfate per
liter of supernatant. The precipitate was collected by
centrifugation at l6,000)<g for 20 min at 20 C, suspended

in distilled water and dissolved by addition of l M NaOH to
a final pH of 7.5. The solution was dialyzed against
several changes of distilled water for 48 h at 4 C prior to
lyophilization. The as- and B-caseinarich fraction was also
solubilized, dialyzed, and Iyophilized.

Preparation of caseinomacropeptide. Caseinomacro—

 

peptide was prepared according to a procedure adapted from

36

Brunner and Thompson (I959). Acid precipitated whole casein
was prepared as for whole casein isolation. After solubi-
lizing in distilled water by addition of I M NaOH to pH 7.0,
one part of rennet was added per l00 parts of Na-caseinate
present in solution at a 3% (w/v) concentration. After
reacting for 60 min at 25 C the pH was adjusted to 4.6 with
l M HCI to precipitate the casein. The precipitate was
removed by centrifugation at l6,000><g for 20 min at 20 C
and trichloroacetic acid added to the supernatant to a final
concentration of 2% (w/v). Insoluble protein was allowed to
settle and the supernatant clarified by filtration. The
filtrate was adjusted to pH 7.0 with l M NaOH and concen-
trated 2 fold by pervaporation. The solution was dialyzed
against several changes of distilled water for 48 h at 4 C

prior to lyophilization.

Chemical Methods

 

Lowry Determination of Protein

 

Protein was determined according to the Lowry modifi-
cation of Markwell _£._l- (I978) using bovine serum albumin
as a standard. Absorbance was read at 660 nm with a Bausch
and Lomb Spectronic 2l. The method employs solubilization
of protein samples in sodium dodecyl sulfate and an increased
copper tartrate concentration to eliminate the necessity of

pretreating membrane or lip0protein samples and minimized

interferences from EDTA and sucrose.

37

Babcock Determination of Fat

 

The Babcock method was used to determine the fat con-
tent of whole milk. The procedure outlined by Atherton and

Newlander (I977) was followed.

Electrophoretic Procedures

 

The following three electrophoretic procedures were
performed using a tube-type apparatus manufactured by

Buchler Instruments. Glass tubes (7.5 cm x 0.60 cm, inner

 

diameter) which were acid-washed and Photo-Flo (0.5%, v/v,
aqueous solution) rinsed were employed. Power wmssupplied by
either a Bio-Rad Laboratories model 400 or a MRA Corporation
model ml58 power supply. Gels were stained for protein with
Coomassie Brilliant Blue R250 according to Weber and Osborn
(I969) using a 2 h staining time and destained with a Bio-Rad

Laboratories model I70 diffusion destainer.

Discontinuous Polyacrylamide Gel Electrophoresis with an

Acid Buffer System

 

Polyacrylamide gel electrophoresis of immunoglobulins
was performed using the acidic buffer system of Reisfeld
_3 _l. (I962). In this system, the basic immunoglobulins
exhibit a reasonable migration distance and degree of resolu-
tion. A running gel with T = 4.l0% and C = 2.60% and a
stacking gel with T = 3.I0% and C = 20.0% were employed.

Chemically-induced polymerization was employed for the

running and stacking gel. Methyl green prepared in stacking

38

gel buffer was added to samples as a marker dye. Electro-

phoresis was performed at a rate of 3 mA/tube.

Discontinuous Polyacrylamide Gel Electrophoresis with an

 

Alkaline Urea—Containing Buffer System

 

Heavy and light polypeptide chains of IgM were elec-
trophoretically resolved using the procedure of Reisfeld and
Small (I966) with these modifications: (a) Heavy and light

polypeptide chains were not separated prior to electropho-

 

resis. (b) The discontinuous buffer system of Davis (I964)

with running and stacking gels containing 8 M urea was

employed. (c) A running gel with T = 4.I0% and C = 2.60% and

a stacking gel with T = 3.I0% and C = 20.0% were employed.

Urea solutions were deionized by stirring with mixed bed ion

exchanger (equal volumes of Dowex 50W—X8 and Dowex AGl-X8,

20-50 mesh) of approximately one-tenth their volume for l2 h

at 4 C. Specific conductivity did not exceed 8-12 umho.
Immunoglobulin M (4 mg) was prepared for analysis by

lyophilizing. After solubilizing in l ml 8 M urea, die

thiothreitol was added to a concentration of 50 mM and

the sample allowed to sit at room temperature overnight.

To maintain the pH, 0.2 ml of pH 8.0 0.5 M Tris-HCI

prepared in urea was added. Iodoacetamide, recrystal-

lized 4)< from distilled water, was added in 0.5 M NaOH to

a final concentration of I50 mM (0.3 ml). After reacting at

room temperature for IS min, the sample was dialyzed 24 h

 

39

against stacking gel buffer at 4 C. Iodoacetamide and all
samples containing iodoacetamide were maintained in the
dark. Electrophoresis was performed at a rate of 3 mA/tube

using bromophenol blue as a marker dye.

Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis
Immunoglobulin M purity was evaluated and molecular
weights of IgM heavy and light polypeptide chains deter—
mined by sodium dodecyl sulfate polyacrylamide gel electro-
phoresis performed according to the general protocol as
described by Weber and Osborn (I969). Protein samples were
incubated at l00 C for l0 min and sucrose added to increase
sample density prior to electrophoresis. Total acrylamide
concentrations (T) of 5.50, 6.50, 7.50, I0.0, l2.0, and l4.0
were employed with a constant cross-linker concentration of
2.64%. A Pharmacia Fine Chemicals low molecular weight
calibration kit was used to prepare standard curves. Asymptot—
ic minimum molecular weights (Segrest and Jackson, I972)
and molecular weights based on retardation coefficients (MW

vs Kr) were calculated (Rodbard and Chrambach, l97l).

Immunological Procedures

 

The following two immunological techniques were
performed using a stabilizing medium of l.0% agarose in
0.0l5 M veronal buffer, pH 8.6, containing 0.02% (w/v) NaN3.
A mold for casting the gel was prepared from 2 glass plates

(8.3 cm x l0.0 cm) from Arthur H. Thomas, C0,, polyethylene

4O

tubing (inner diameter l.l4 mm, outer diameter l.57 mm) from
Clay Adams, and metal pinch clamps. One plate was coated
with silicone vacuum grease for easy removal after the
agarose had hardened and clamps removed. A gel of nominal
thickness l.5 mm was obtained. Anti-sera and sample wells
of 3 mm diameter were cut with a sharp-edged stainless—steel
tube. Agarose plugs were removed with a Pasteur pipette
connected to a water aspirator. Wells were filled with

either 7 ul of sample or anti-sera. Plates were allowed to

 

develop in a closed,leveled chamber containing vessels of
water and thymol crystals.

Following development the plates were pressed, washed
in phosphate buffered saline for 24-48 h and stained
with Coomassie Brilliant Blue R250 following the procedure

of Weeke (I973).

Two-Dimensional Double Diffusion

 

A procedure adapted from Ouchterlony (I949) was used
for two-dimensional double diffusion immunological identifi—
cation of IgM. An electrophoretically homogeneous sample
was subjected to two-dimensional double diffusion versus
rabbit anti-bovine IgM (u-chain specific) anti-sera. Prior
to analysis the sample was dialyzed against 0.0l5 M veronal
buffer, pH 8.6, for 8 h at room temperature. In preparing
wells, a template with 6 antigen and 6 anti-sera wells placed
on intersecting lines and with corresponding well-centers

spaced from 0.50 to l.40 cm apart was utilized. Diffusion

41

was allowed to proceed for 90-I00 h.

Two-Dimensional Single Radial Diffusion

 

Two-dimensional single radial diffusion performed accor-

ding to a modified procedure of Mancini g1 g1. (I965) was
employed for IgM quantitation in whey. Antibody concentra-

tions of 0.l2-0.60% (v/v) and IgM concentrations of 0.03-
0.50 mg/ml were used. The concentration of IgM in standard

solutions was based on absorbance at 280 nm and a specific

 

absorption coefficient (Eiéo nm) of l2.6 (Mukkur and Froese,
l97l). Prior to application,the IgM solution was dialyzed
against 0.0l5 M veronal, pH 8.6, buffer for 8 h at room
temperature. Whey samples were concentrated approximately
2)< by dialyzing samples against a I5-20% (w/v) polyvinyl-
pyrrolidone solution prepared in 0.0l5 M veronal, pH 8.6,
buffer at 4 C. Diffusion was allowed to proceed for 90-
l00 h. After staining, precipitate diameters were measured
using a Nikon model 6 Shadowgraph Microcomparator equipped
with an Ehrenreich Photo-Optical Industries Electromike.
Plots of precipitate diameter versus log (concentration)

yielded straight lines over limited concentration ranges.

Creaming-Studies Associated Procedures

 

Preparation of Samgles

 

Milk. Creaming studies were generally performed using
fresh whole milk. For some studies l-2 day old milk stored

at 4 C was employed. Prior to use all milk was placed in a

 

42

water bath at 55 C for 30 min with occasional gentle mixing.
If skim milk was required, either cold- or warm-separated
whole milk was centrifuged at l,000)<g for 20 min at room
temperature. Skim milk was removed with a syringe and
needle inserted through the compacted cream layer. When
required, washed milk fat globules were added to give a fat
content of 3.2—3.6%. Substances added to milk were prepared
in SMU or dialyzed against several changes of SMU for 6-8 h

at room temperature.

 

Heated milk. Milk, either whole or skim, devoid of

 

creaming capacity due to inactivation of the heat-labile
factor was prepared by placing milk in a closed container in
a water bath at 75 C for 30 min with occasional gentle
mixing.

Homogenized milk. Skim milk devoid of creaming

 

capacity due to inactivation of the homogenization—labile
factor was prepared by homogenizing skim milk (40 C) at
2500 psi in a single stage model C-8 C.W. Logeman Co.
homogenizer. If required, washed milk fat globules were
added to give a fat content of 3.2-3.6%.

Washed milk fat globules. Washed milk fat globules

 

were prepared from whole milk previously warmed to 55 C for
30 min. A cream layer was obtained by centrifuging the
whole milk at I,000><g for 20 min at room temperature.
After removing the skim layer, four volumes of distilled

water at 40 C was mixed with the cream. The solution was

43

centrifuged again and the underlayer removed. This process
was repeated 2 additional times.

Synthetic milk fat globules. Synthetic fat globules,

 

or stable fat globules prepared by emulsification in an
aqueous protein solution, were prepared using butter fat
and either K-casein, an as- and B-casein rich fraction, or

a mixture of these 2 samples (65—, B-caseinzk-casein::l5:85,

w/w). Butter oil obtained from churned washed milk fat

globules was washed twice by ultracentrifugal separation
(l,000x g) for IS min at room temperature) using I M
NaCl solution. Protein solutions at a 0.35% (w/v)
concentration were prepared in SMU. A coarse emulsion
was formed by mixing 4 g of butter oil (55 C) with l00 ml of
protein solution (55 C) in a Waring blender for 45-60 sec.
After sitting at 55 C for l0—l5 min to reduce foam, the
mixture was homogenized at approximately 500 psi with a
C.W. Logeman Co. hand homogenizer. The fat globules were
collected by centrifuging at l,000x g for 20 min at room
temperature. They were then washed 2 times with 4 volumes
of 55 C distilled water.

Wheniabserved microscopically (I000 x), the fat globule size

distribution was predominantly in the 2-511diameter range.

Quantitation of Creaminngapacigy

 

Cream volume. Cream volume was determined by placing

 

37 C samples in I0 ml graduated cylinders into an 9-Il C

 

44

water bath. The water depth was sufficient to cover the

sample. The milliliters<rfcream layerfbrmea was read directly
to the nearest 0.051nlfron1the graduated cylinder scale with the
aid of transmitted light. CWeam volumesvmre generally'read at
I, 2, 3, 4 and 24 h intervals. Afterlllithe graduated cylin-
der was placed in a 4 C cold room. If a distinct cream line
had not developed but clusters were visible throughout the

cylinder, a cream volume of l0.0 was reported.

Cluster time. The cluster time, or the time in minutes

 

 

for visible clusters to form, was also used as a measure of
creaming capacity. A procedure similar to that employed by
Dunkley and Sommer (l944) was used. Creaming cells with
inside dimensions 76 mm x 20 mm x I mm were constructed

from thin microscope slides. Two cells were mounted on a
thick glass base plate. The cluster time was determined by
placing 37 C samples in the creaming cells into a plexiglass
water bath maintained at 9.5 - I0.5 C. The milk was con-
tinuously observed using transmitted light. The time at
which clusters were first visible was reported as the cluster

time.

Enzymic Treatment of Milk Fat Globules

 

Washed milk fat globules were treated with various
enzymes in order to examine the role of membrane constitu-
ents in creaming. Twenty milliliters of washed milk fat

globules was added to l00 ml of reaction buffer containing

45

enzyme to give a fat content of approximately 6%. Controls
containing no enzyme were run in each case. For trypsin a
reaction medium of 0.05 M Tris-HCI, pH 7.8, a reaction time
of 20 min at 37 C, and a buffer enzyme concentration of

50 ug/ml wemaemployed. For pronase the same reaction condi-
tions were used except for a buffer enzyme concentration of
ISO ug/ml. For mixed glycosidases (content and activities
are given in the Appendix, Table A3) a reaction medium of
0.l0 M Na-citrate, pH 4.6, a reaction time of I2 h at 37 C,
and a buffer enzyme concentration of I mg/ml was employed.
For neuraminidase the same reaction conditions were used
except for a buffer pH of 6.0 and enzyme concentration of
IOO ug/ml. For the latter 2 treatments, I drop of toluene
was added to inhibit bacterial growth. For samples treated
with mixed glycosidases and neuraminidase after trypsin
treatment, appropriate buffers were employed using the
washing procedure given below between treatments. For milk
fat globules treated with mixed glycosidases and neuramini-
dase simultaneously, the Na-citrate, pH 6.0, buffer was
employed using the enzyme concentrations listed above. The
neuraminidase preparation exhibited no protease activity as
measured with the Azocoll (Calbiochem-Behring Corp.) assay
at pH 6.0 and 37 C. The mixed glycosidase preparation
exhibited very low protease activity after a reaction time
of 4.5 h at pH 4.6 and 37 C. A 3-fold weight excess of

soybean trypsin inhibitor (l mg inhibits l.5 mg trypsin)

 

46

was added to trypsin treated milk fat globules to stop
digestion. Milk fat globules were collected by centrifu-
ging at l,000x g for 20 min at room temperature. They were
then washed 3 times with ID volumes of 40 C distilled water
using the same conditions for separation. Milk fat glo-
bules were added back to skim milk to a fat content of

3.5% for creaming studies.

Spectrophotometric Evaluation of IgM Cold-Induced Aggre-

 

gation

A procedure similar to that of Payens (I968) was
employed. Immunoglobulin M purified as previously de-
scribed was concentrated to approximately l.2 mg/ml using a
Millipore Corp. Immersible CX unit (nominal molecular weight
cut-off l0,000 daltons). After concentration the solution
was dialyzed against 3 changes of SMU for 6 h at room tem-
perature. The sample absorbance was followed at 344 nm
using a Gilford Instrument Laboratories, Inc. model 240
Spectrophotometer equipped with a model 245lA Automatic
Cuvette Positioner while decreasing or increasing its tem-
perature by means of circulating water at 0 C or 40 C. The
temperature in a control cuvette was read intermittently. A
final sample temperature of either l0 C or 37 C was achieved.
The absorbance of an IgG and void volume fraction was also

followed after dialysis against SMU.

RESULTS AND DISCUSSION

Isolation and Identification of Immunoglobulin M

 

Isolation

 

Immunoglobulin M (IgM) has been prescribed a role in
milk fat globule (MFG) clustering due to its ability to
restore creaming to heated milk (Franzen, I97l) or to bring
about clustering of MFG in model systems (Gammack and Gupta,
l970; Payens and Both, I970). To confirm these observa-
tions it was necessary to obtain an IgM fraction from cows'
milk, colostrum, or blood serum. Salt fractionation and
gel filtration were employed as outlined in Figure I to
isolate IgM from milk.

Lipoproteins and residual lipid material were removed
from the crude immunoglobulin fraction using ultracentri-
fugation, to avoid impairing column performance and to
reduce the size of the void volume fraction. Table I de-

tails the relative purification and recovery of IgM (con—

sidering whey as l00%) for 3 steps in the isolation scheme.
When the crude immunoglobulin fraction was subjected to gel
filtration on columns of Bio-Gel A-5m and A-0.5m connected

in tandem, 3 prominent peaks were obtained (Figure 2). The

Bio-Gel A-0.5m column was employed to increase resolution

47

 

48

 

mewmposaoawF 1
cwrsnoFmocsssw

 

N cw em.o mF.m com wvzsu
:wpsno_moczeew
0 mm mm._ m.m~ mmm mussu
:wrzno_m
m we «v.0 w_.w oom~ wussu
- oo_ Aeo.o e.w_ omeme soez
Axv Aee\msv Are\mev
cowpmowewgzm >sm>oomm cowpmgpcmocou cowpasucmocou AFEV
cred ZmH zmH :wwposa mE:_o> :owuuwsd

 

.mEmcom cowumPOm? esp cw mawpm empomrmm pm ZmH to cowpmowwwcza ucm zcm>oows one ._ wrnmp

Figure 2.

49

Gel filtration chromatogram of a crude immuno-
globulin preparation. Bio-Gel A-5m (2.6 cm x
65 cm) and A-0.5m columns connected in tandem
were employed. Approximately 500 mg of protein

containing 40 mg of IgM was applied to the
column.

50

 

N 0.33....

20:043.".
8 8

SF 2: 3 8

 

8.:

8m

 

cc.—

 

 

 

 

5l

between the latter 2 peaks. The eluted fractions contain-
ing the first peak, corresponding to the void volume, was
translucent. The second peak was identified as IgM and
the last peak as primarily 198 using electrophoretic and
immunological analysis (Results presented in the sections
Electrophoretic Analyses and Two-Deminsional Double Diffu-
sion). An IgM sample was obtained by pooling fractions

44-55. Sixty percent of the IgM applied to the column was

 

recovered in this fraction. The remainder was distributed
over the next l5 fractions which also contained IgG.
Re-chromatography of the IgM sample after concentration
to 3.6 mg/ml yielded the chromatogram presented as Figure3.
The sample was essentially homogeneous except for a small
void volume peak, probably consisting of aggregated IgM.
The recovery of IgM was 65%. The significant loss of IgM is
attributed to the removal by centrifugation (27,000><g) for
30 min) and filtration of protein aggregates (turbidity)
which formed during storage at 4 C and concentration.
Protein aggregation is characteristic of purified IgM prepa-
rations (Metzger, I970). As re-chromatography did not
contribute to an increase in IgM purity, it Was not included

in the general isolation scheme.

Electrophoretic Analysis
Electrophoretic patterns of acid whey, crude globulin,

crude immunoglobulin (after removal of lipoproteins), and

52

Figure 3. Gel filtration chromatogram of IgM. Bio-Gel A-5m
(2.6 cm x 65 cm) and A-0.5m (2.6 cm x 65 cm)
columns connected in tandem were employed.

Approximately 20 mg of IgM was applied to the
column.

53

0.60

280

0.40

 

0.20

 

 

20 40 5"
FRACTION

Figure 3

54

pooled fractions obtained from gel filtration of the crude
immunoglobulin preparation (Figure 2) are presented in
Figure 4 . The basic immunoglobulins separate into
distinct zones with the acid buffer system employed. The
immunoglobulins form diffuse zones near the stacking/run-
ning gel interface when alkaline buffer systems are used
(Melachouris, I969; Franzen, l970). An enrichment of
immunoglobulins is noted on progression from the whey to
the crude globulin and crude immunoglobulin samples.
Fractions 44-54, corresponding to the second peak in the
gel filtration pattern, contained only IgM. Fractions
55-72 contained 196 in addition to IgM. Fractions 73—l05
contained primarily IgG. The void volume fractions failed
to yield protein bands which stain with Coomassie Brilliant
Blue R250. IgA, which is commonly absent from cows' milk
(Kumar and Mikolajcik, I972), was not present in the
samples.

When subjected to reduction and alkylation, electro-
phoretically homogeneous IgM yielded numerous components
which were resolved by electrophoresis in an alkaline urea-
containing buffer system (Figure 4, gel 8). Immunoglobulin
heterogeneity or numerous heavy and light chains are
expected in polyclonal antigenic responses (Kabat, I976).
The photographic technique employed here did not permit
discernment of all protein bands which actually appear as

numerous diffuse zones.

Figure 4.

55

Electrophoretic patterns characterizing the iso-
lation of IgM from milk. Samples applied to the
gels are whey (I), crude globulins (2), crude
immunoglobulins (3), and fractions obtained by
gel chromatographic isolation of the crude
immunoglobulin fraction (see Figure 2): fractions
44-54 (4), fractions 55-72 (5), fractions 63-78
(6), and fractions 88-l05 (7). Gel 8 represents
alkylated, 2-mercaptoethanol-reduced IgM. Dis-
continuous polyacrylamide gel electrophoresis
with an acid buffer system was employed for

gels l-7 and an alkaline, urea-containing buffer
for gel 8. Protein loads on the gels were:

90 ug (l), 40 ug (2-7), and 200 pg (8).

IgGl _
I962-
BSA-
a-Ia—

B'lg-

 

 

56

 

 

Figure 4

57

Sodium dodecyl sulfate polyacrylamide gel electrophero-
grams of IgM are presented in Figure 5. As expected, 2
major protein zones representing heavy and light chains
were obtained. A faint band having a mobility slightly
greater than the heavy chain was evident at each gel con-
centration. Table 2 lists the apparent molecular weights
of heavy and light chains for each gel concentration. The
observed molecular weight for heavy chains varied with

acrylamide concentration while for light chains was inde-

 

pendent of acrylamide concentration. This is attributed
to the relatively high carbohydrate content of heavy chains
(l2.0%) and low carbohydrate content of light chains (less
than 0.5%) (Beale and Buttress, I972; Segrest and Jackson,
I972). Asymptotic minimum molecular weights of 74,500 and
27,500 were estimated for the heavy and light chains,
respectively. From a plot of molecular weight versus
retardation coefficient (Figure 6), molecular weights of
70,000 and 25,000 were obtained for the heavy and light
chains, respectively. Mukkur and Froese (I970), employing
gel filtration, reported values of 76,000 and 22,500 and
Beale and Buttress (I972), employing sedimentation equili-
brium, reported values of 6l,800 and 22,800 for heavy and

light chain molecular weights, respectively.

TWO-Dimensional Double Diffusion

 

The two-dimensional double diffusion pattern of pooled

fractions 44-55 (Figure 2) developed against anti-IgM

 

Figure 5.

58

Sodium dodecyl sulfate polyacrylamide gel elec-
tropherograms of molecular weight standard
proteins and IgM. Gel pairs I, 2, 3, 4, 5,

and 6 contained total acrylamide concentrations
of 5.50, 6.50, 7.50, I0.0, l2.5 and l4.0%,
respectively. IgM gels were charged with

25 ug of protein.

59

 

"In" 5

6O

 

.mpcmeszmmmE m-m to mmmsm>m asp mew; mmwuwfivnos m>eumpmmm

 

oom.mm
oom.¢n

ooo.~m ooo.wm ooo.wm ooo.nm oooawm oooanm :mmzo pcmwb
coo.mm ooo.¢m ooo.mm ooo.mw ooo.¢w ooo.©w :wmco x>mmI

 

unmwmz Lm_:um_oz
Essecwz oep0pa5>m<

o.¢~ m.- o.o_ om.~ om.o om.m

 

Axv mcoepmcpcmucou Pow to
cowuoczd m mm pgmwmz smpsom—oz :wmuosa

 

.cowpwcpcmocoo Pom to :owuoczw m we mwmmgozaosuowpm pom mvwEm—»L0m>P0Q mpmwpzm
onmvou Eavuom >3 umcwEmemu mm mcwmco pcmwfi vcm >>mmg zmH mo mpsmwmz Lm_:ompoz .N mFQmH

10.0

2.0

Figure 6.

6l

 

0.02 0.06 0.10

Plot of molecular weight versus retardation
coefficient derived from sodium dodecyl sulfate
polyacrylamide gel electrophoresis data (see
Figure 5). Arrows indicate 'the retardation
coefficients for IgM heavy (A) and light (B)
chains.

62

anti—serum (u-chain specific) is presented in Figure 7A.
The precipitin reaction verifies that the designated sample
is IgM. A faint reaction was obtained in the first set of
wells versus anti-IgG anti-serum (y—chain specific), see
Figure 7B. The faint reaction demonstrates that an IgG
contaminant is present in the sample - but at a very low
concentration. Based on chromatographic, electrophoretic,
and immunological results, fractions corresponding to 55-66

(Figure 2) were collected for experiments requiring IgM.

Confirmation of Immunoglobulin M

 

as a Creaming-Active Component

 

Effect on Creaming of Heated Milk

 

Franzen (l97l) demonstrated that IgM is the euglobulin
component which restores creaming to heated milk. These
results are confirmed by the data presented in Table 3.
Whereas IgG and the void volume fraction failed to restore
creaming to heated milk, IgM did restore creaming.
Increased levels of IgM were found to restore increased

levels of creaming capacity.

Identification of Immunoglobulin M as a Creaming-Active

Component in Raw Milk

 

Removal from the reaction with specific anti-serum.

 

To further examine or verify the role IgM plays in the

creaming phenomenon, antibodies specific for IgM, IgG,

 

63

Figure 7. Two-dimensional double diffusion patterns of
pooled gel filtration fractions 44-55 (see
Figure 2) developed against (A) anti-IgM anti-
serum (u-chain specific), and (B) anti—IgG
anti-serum (y-chain specific).

 

64

anti-IgM

anti-IgG

F
lgute 7

65

.momxpoco opoowpooo mo omoso><

n

.xree 30» oco oopoo: mg“ op ooooo mo: 22m
.xrws oopoo; to mpsoo m op ooooo 003 cowporom :zmizmH so .oE:Fo> owo> .me mo peoo ocoo

 

 

 

0.e 00._ 00._ 0N._ 00., s_ae 300
0.0 00._ 00._ me._ 00._ _e\0e 00.0 .200+
0.0 00._ 0N._ 0N._ 0N.F _s\0e 0_.0 .200+
0.0 0_._ 0N._ 0N._ 00.0 Pe\0e 00.0 .200+
0.3 00.0 00.0 00.0 00.0 2.00. 000.0 .207.
+ 00 00.0 0_.0 00.0 00.0 osspos 0000+
+ 00 00.0 0_.0 00.0 00.0 _E\0e 0.F .00H+
+ 00 00.0 0_.0 00.0 00.0 _e\0e 0.F .000+
+ 00 00.0 0F.0 00.0 00.0 _e\05 0.0 .00H+
spas 00000:
A0000 0 00 e 0 e N e _
meH prmzru A—E o_.\FEV mE:—o> Emeo mwFQEmm

o

a

 

.x_0e 00000;

to mcwsooso co mcowuoose ZmH oco .oE:_o> owo> .omH ooquome mo powwow one .m oFoop

66

and IgA heavy chains were added to raw milk. Indeed, if

IgM is involved, the interaction with antibodies should
remove it from the reaction, and decrease the extent of
creaming. The results presented in Table 4 indicate that
only in the case of added anti-IgM is the creaming effected,
as measured by fat globule clustering time. At low levels
of added anti-IgM, it is likely that soluble complexes
involving only part of the IgM are formed due to an antigen
excess. Only when the level is increased to amounts which
bind significant levels of IgM is an effect seen.

Removal from the reaction with 2-mercaptoethanol.
Mercaptoethanol (ME) treatment has been used as a means of
differentiating between certain classes of immunoglobulins
(Scott and Gershon, I970; Kabat, I976). Whereas ME-sensi-
tive immunoglobulins lose their antibody combining activity,
ME-resistant immunoglobulins suffer no or slightly diminished
antibody combining activity. Although exceptions have been
reported (Adler, I965; Kim 33 gl., I966), ME—sensitive
immunoglobulins are generally l9S-IgM, while ME-resistant
immunoglobulins are usually 7S-IgG molecules. Since ME
treatment destroys the antibody activity of bovine IgM
molecules (Butler, I969), its addition to milk may be
expected to eliminate fat globule clustering and creaming.
Data reported by Franzen (I97l), presented in Table 5,
demonstrate that creaming behavior of milk is eliminated

upon the addition of ME.

67

Table 4. The effect of IgM, IgG, and IgA specific anti-
sera on creaming of milk.

 

 

Sample Cluster Timea (min)
Raw milk 6.7

+ 0.025 ml anti-IgM/2 ml 6.5

+ 0.05 ml anti-IgM/2 ml 6.5

+ 0.20 ml anti-IgM/2 ml 9.0

+ 0.30 ml anti-IgM/2 ml 2l.3

+ 0.30 ml anti-IgG/2 ml 6.8

+ 0.30 ml anti-IgA/2 ml 6.7

 

a
Average of duplicate analyses.

Table 5. The effect of 2-mercaptoethanol on creaming of

 

 

 

milk.a
Sample Cream Volume (ml/l0 ml)
l h 2 h 24 h
Raw milk 0.00 3.20 I.l5
+ 0.5% 2-mercaptoethanol 0.00 3.50 l.20
+ 2.5% 2-mercaptoethanol 0.00 0.00 0.l0

 

aAdapted from Franzen (I97l).

68

At the lower level (0.5% ME), other milk proteins may have
been preferentially reduced. However, at the higher level
(2.5%), creaming was definitely eliminated. While indi-
cating that IgM is involved in creaming, these results do
not eliminate an indirect effect due to reduction of other
proteins or components in milk. The failure of alkylated,
reduced IgM to restore creaming to heated milk supports
the notion that creaming is eliminated due to the effect
on IgM (Franzen, l97l).

Effect of adding immunoglobulin M to raw milk. The

 

addition of IgM to raw milk would be expected to augment
creaming if it is involved in the phenomenon. The data of
Table 6 support the contention that IgM is involved in

creaming or fat globule clustering.

Immunoglobulin M Cryoaggregation

 

After it was established that IgM is involved in fat
globule clustering, its cryoglobulin character, i.e., rever-
sible low temperature-induced insolubility, was examined.
The studies of Payens (I968) and Rhee (I968) with euglobulin
and Payens and Both (I970) and Franzen (l97l) with IgM
implicated this characteristic of the protein as essential
to fat globule clustering. The proteins were designated
as cryoglobulins based on temperature-dependent turbidity

measurements or sedimentation properties. The authors

69

.momAF0oo ouoow—ooo mo ouoso><

 

 

2::

v

n

o
.XPWE to mpsoo 0 op ooooo 003 22m so cowp:_om 32mizmH mo “Loo oco Locpwmo
0.0 00._ 00._ 0A._ 00._ _E\0E 00.0 .200 +
_.0 mm._ om._ mm.~ mm._ xF05 30¢
; em 5 m c N g _

 

meek Lopmoro

APE op\_sv noE3Fo> Eoocu III. ooposom

.xpws do mowsooso co EmH mo poomwo och .o opooh

70

failed to examine or address the evidence that immunoglo-
bulins, and in particular IgA and IgM, polymerize or aggre-
gate when purified (Butler, l969; Metzger, l970).

According to Zinneman (I978) human cryoglobulins may be
classified into 3 categories based on their heterogeneity:
(a) single monoclonal immunoglobulins, (b) mixed cryo-
globulins, consisting of one monoclonal and one polyclonal
immunoglobulin, and (c) mixed cryoglobulins, consisting of
only polyclonal components. Mixed cryoglobulins are formed
by two immunoglobulins, neither of which is a cryoglobulin
in the single state. Data were presented to show that most
cryoglobulins fall into the third category (I86 of 4ll
cases) with the majority of these consisting of IgM and IgG
(I68 of I86 cases). It was suggested that IgM may combine
with IgG bound to an unknown antigen. Brouet _g _1. (I974)
reported similar results in a study of 86 cases of cryoglo-
bulinemia - 50% of the cases were assigned to the third
category with 85% of the cases involving IgM and IgG. Both
authors found the remainder of the cases in group 3 to
involve complexes containing IgM, IgG, and IgA. These
studies demonstrated that a single immunoglobulin with the
property of reversible cryoprecipitation is generally mono-
clonal. Based on experiments which demonstrate a role for
IgM, and not IgG or IgA, in fat globule clustering and the
electrophoretic pattern of alkylated, reduced IgM the
cryoglobulins in milk fall into the third category.

7l

The presence of cryoglobulins in serum is commonly
associated with clinical symptoms characteristic of immuno-
proliferative and autoimmuno disorders. In approximately
l0% of the cases, acute and severe symptoms necessitate
emergency treatment with plasmapheresis and chemotherapy.
More than half of the patients with high levels of single
monoclonal cryoglobulins and approximately l5% of those
with mixed polyclonal cryoglobulins are asymptomatic
(Brouet g; _l., I974). Cows fail to show symptoms of
cryoglobulinemia.

It is of interest that Franzen (l97l) found that the
reduction of IgM inhibited creaming (Table 5) while Payens
and Both (I970) reported that reducing agents did not inhibit
IgM cryoaggregation. If the same functional groups are
responsible for cryoaggregation of IgM and fat globule
clustering, as suggested by Payens and Both, one would
expect parallel results from the above experiments.

The following experiment was performed to examine the
cryoglobulin character of isolated IgM. By studying
conditions which influence cryoaggregation of IgM, factors
which influence fat globule clustering should become
apparent and better understood if the presently accepted
theory is correct. A solution of IgM was concentrated
to approximately l.2 mg/ml and its turbidity monitored
while temperature was varied from l0-37 C. The absorbance

of IgG (l.l mg/ml) and the void volume fraction (diluted

72

I:l with SMU) was also monitored. Results are presented

in Table 7. Absorbance of the IgM solution increased
slightly when temperature was decreased, and decreased

as temperature was increased. This was also observed for
the void volume fraction. Absorbance of the IgG fraction
did not change as the temperature was varied. During
concentration of IgM at room temperature in preparation

for this experiment, the absorbance at 280 nm increased
approximately 2.6 fold while absorbance at 344 nm increased
approximately ll fold. The significant increase in absor—
bance at 344 nm which was observed during concentration

and the slight increase recorded as a result of wide
fluctuation in temperature indicate that a concentration
and purity dependent aggregation may be the more significant
factor promoting protein aggregation. As mentioned pre-
viously, this property is characteristic of IgM and IgA
(Butler, I969; Metzger, l970). Based on the above discus-
sion and experiment it was decided to further examine the

role of IgM in creaming.

Quantitation of Immunoglobulin M

 

Participating in Creamigg

 

Experiments were performed to determine exactly how
much of the IgM in milk participates in a single creaming.

Milk was held quiescently overnight at 4 C and then

73

 

 

 

E: com um oocoosomo< o_osom

mom._ moo.o mwo.o 0.0m

mw¢.~ woo.o mm_.o o.o~

omm.~ woo.o mwo.o 0.0m

mm¢._ woo.o mmF.o o.o_

mF¢.— woo.o mo—.o m.mm
wszpo> Uwo> me zmH

Auv osspogoosoh

 

.cowpoocm

osspo> owo> one 0:0 .me .ZmH to cowmsoomeo ogw :o ogoposooeop to poomwo one

.0 wFQmH

74

centrifuged for ID min at I000x g. The skim milk layer,
which will be referred to as gravity-separated skim milk,
was examined for IgM. Approximately 93% of the IgM
(0.0689 mg/ml) remained relative to a warm-separated
control which contained 0.0743 mg/ml.

To determine if the amount of IgM associated with the
milk fat globules (MFG) could be increased, known amounts
of IgM were added to raw milk which was allowed to cream
overnight. The amount of IgM remaining in the gravity-
separated skim milk was determined as above. Results are
presented in Table 8. The calculated IgM concentration is
the amount in warm-separated skim milk (0.05l3 mg/ml)
plus the amount added to the original raw milk. Based
on the ratio of calculated to experimental results or the
% recovery, it was apparent that the added IgM, or some
fraction of the added IgM, did not associate with the MFG.
If the reaction merely involved aggregation of IgM and a
non-specific precipitation onto the MFG, one would expect
the added IgM or some fraction thereof to be associated
with the MFG. These results are consistent with the
hypothesis that there are a limited number of binding

sites capable of reacting or interacting with IgM.

.0000F000 opoow_ooo so omoso><0

 

75

 

 

oo_ oo._ mFP.o m__.o
FOP mom.o wmmo.o mmmo.o
m__ nmm.o mmmo.o omuo.o
mo_ omm.o Fmoo.o Fmoo.o
om_ 000.0 mmoo.o wmmo.o
FNF mom.o mmmo.o mmmo.o
opoucoswsooxm oopopoopou
0Lo>ooom & .oxm\.oFou ovpom

APE\mEv :owpmsgcoocou zmH

 

.zmH sue: oopcoeo_ooom xpws mo mcwaooso cw oo>_o>:w zmH wo cowpouepcooo .w oFDoH

76

Examination of the Ability_of Gravity:Sgparated

 

Skim Milk to Sgpgort Creaming

 

Creaming of Recombined Milk-Washed Milk Fat Globules and

 

Gravigy—Separated Skim Milk

 

Washed MFG were added to gravity-separated skim milk to
determine if the remaining IgM is capable of supporting
creaming. As indicated in Table 9, the recombined milk
failed to cream. This indicated that only a fraction of
the IgM in milk is ”active" or capable of promoting fat
globule clustering, or that another component in the process
is limiting and absent from gravity-separated skim milk.

In order to examine these alternatives, the following

experiments were performed.

Creaming of Recombined Milk- Washed Milk Fat Globules and

 

Gravity-Separated Skim Milk from Raw Milk Supplemented with

 

Immunoglobulin M

 

The following experiment was performed to determine if
the IgM remaining in gravity-separated skim milk is
"inactive". Raw milk was supplemented with IgM (0.25 mg/
ml) and the resulting gravity-separated skim milk examined
for its ability to support creaming upon addition of washed
MFG. Since the added IgM will not be actively involved
in creaming and will remain in the gravity-separated skim
milk (Table 8), the gravity-separated skim milk would be

expected to support creaming on the addition of washed MFG.

77

.mcwst pcoupwssopCW
£003 :05 om Low 0 mm op oessoz coon 0:0 oosooso 00; gown: xpws 30; 0o oopmwmcoo Pospcouo

.mo000000 op0o0_ooo 0o omoco><o

.N0.0 00 0000000 000 _0000 0 00 00000 0003 00:0

 

 

 

0.__ .00._ 00.0 00.0 00.0 o_000000
+ co om.o or.o o—.o o~.o mm: + vam xwm>wgw
—.op ow.~ om.© mo.m om.m x—wE 36m
0 0N 0 0 0 N 0 _ ‘1
Acwsv ooewp Loum:_u A—E o_\_ev ooaopo> Eoosu oo_osom

 

.Aoazv mo_000_0 000 s_0s 000003 to 00000000
ocu mcezoFFom mcwsooso psoooom op xF0E 50x0 ooposooomixuw>0em to cowpoopo>m .m o_ooh

78

Data presented in Table I0 indicate that the gravity—
separated skim milk failed to support creaming, favoring
the alternative hypothesis that gravity-separated skim
milk fails to support creaming due to the absence of a

limiting component other than IgM.

Effect of Immunoglobulin M Isolated from Gravity-Separated

 

Skim Milk on Creaming of Heated Milk

 

To further examine the possibility that gravity-
separated skim milk fails to support creaming due to the
absence of another limiting component, the IgM remaining
in gravity-separated skim milk was isolated according to
the procedure outlined in Figure I and its ability to
restore creaming to heated milk was examined. Results are
presented in Table II. "Inactive" refers to IgM isolated
from gravity-separated skim milk which failed to support
creaming and "total” refers to IgM isolated from warm-
separated skim milk. Results are sufficiently similar to
conclude that there is no significant difference in the
ability of the two IgM fractions to restore creaming to
heated milk, indicating further that another limiting com-
ponent is absent from the gravity-separated skim milk.
Since it is not the heat-labile component, one could specu-
late that it is the homogenization-labile component ori-

ginally referred to by Samuelsson e5 gl. (l954).

79

.00033000 opoowfioo0 0o 000Lo><0

.00.0 00 0000000 000 _0000 0 00 00000

 

 

 

AFE o_\_EV nos:_o> Eoogu

0Lo3 an: .xrvs 30; 0o 00000 m op 00000 003 :20 Lo cowpzpo0 DthzmH mo 0000 oco Locpvmm
0.0m 00.0 00.0 00.0 00.0 002 + AzmH+v 5030 >00>0Lw
+ 00 00.0 0N.0 00.0 00.0 003 + 0000 000>0e0
o.m 00.0 oo._ ON._ om._ Fs\ms 0N.0 .ZmH +
_.N mm._ 00.0 mm._ mm._ xFWE 30m
0 0m 0 m 0 N c _
A00EV 00500 Lop0zfiu 0o_os0m

 

.Awmzv 0opoooFm p00 xF0E 000003 mo :owpw000 asp mcw3oFFom mavsoogo ugooas0
op 3002 30s 0mucoso_oa:0-ZmH Eos0 3005 5030 0op0sooo0ixuw>0gm mo cowpoo~0>m .oF o_000

 

8O

.000>_0:0 000000000 0o 0m000><0

 

 

 

 

.0005 0o 00000 0 o0 00000 003 :20 0o :o000000 020-200 0o 0000 00o 0000000
0.0 00.0 0N.0 00.0 00.0 0000 300
0.0 00.0 00.0 00.0 00.0 PE\0E 0N.0 .000 =00000. +
0.0 00.0 00.0 00.0 00.0 _e\00 00.0 .000 .00000s +
0.00 00.0 0N.0 00.0 00.0 00000 00.0 .300 .00000= +
0.0 00.0 00.0 00.0 00.0 _e\00 0N.0 .200 .0>000000= +
0.0 00.0 00.0 00.0 00.0 _E\00 00.0 .200 .03000000. +
0.00 00.0 00.0 00.0 00.0 Pe\0e 00.0 .200 =0>000000= +
+ 00 0N.0 00.0 00.0 00.0 0000 00000:
0 0N 0 0 0 N 0 _
A0050 00500 0000000 APE o_\FEv 0E:Fo> E0000 00F0E0m
.xFWE 000000 0o 00050000 co xF0E E000 . =0>00o000= . 000000000
-000>00m 0o 1 000000: . 00000000010003 Eo0w 0000Fo00 :00 0o 000000 000 .00 00000

8l

Examination of the Role of the Antigen-Binding

 

Properties of Immunoglobulin M in Creaming

 

Isolation of Fab Fragments

 

Tryptic digestion and gel filtration were employed to
obtain Fab fragments from IgM. The gel filtration chro-
matogram of digested IgM-7S subunits is shown in Figure 8.
Fractions 53-58 were collected as the Fab-containing frac-
tion. The peak maximum corresponded to a molecular weight
of 44,700. Sodium dodecyl sulfate polyacrylamide gel
electropherograms (Figure 9) indicated the presence of low
molecular weight contaminants. Species corresponding to
intact heavy chains were not present. Light chain and Fd
fragment molecular weights were 27,900 and 33,000, respec-

tively.

The Effect of Fab Fragments on Creaming

 

If the IgM-MFG interaction is specific and involves
the antigen-binding portion of the molecule, the addition
of Fab fragments to milk should inhibit creaming by binding
with reactive sites (antigens) without promoting cross-
linking or clustering. When Fab fragments dispersed in SMU
were added to raw milk (final concentration 0.I0 mg/ml) the
results shown in Table l2 were obtained. These data impli-
cate a specific interaction in which the antigen-binding
portion of the IgM molecule participates. An antigen-

antibody interaction mode is suggested as opposed to an

82

Figure 8. Gel filtration chromatogram of alkylated, 2-
mercaptoethanol—reduced, trypsin-treated IgM. A
column of Sephacryl S-200 (2.6 cm x 90 cm) was
employed.

83

8—

 

a 00:3“.

ZO_._.U<~_ ”—
S

 

 

3

cm

 

 

 

 

2...

8m

am...

an...

Figure 9.

84

Sodium dodecyl sulfate polyacrylamide gel electro-
pherograms of Fab fragments — pooled gel filtra-
tion fractions 55-58, see Figure 8 - (I and 2)

and IgM (3). Gels l and 3 contained 25 ug of
protein and gel 2 contained 45 pg of protein.

 

Figure 9

86

Table l2. The effect of IgM-derived Fab fragments on
creaming of milk.

 

 

Samplea Cluster Timeb (min)
Raw milk 8.0
+Fab fragments l2.5

 

aEither one part of Fab fragment-SMU solution or SMU was
added to 3 parts of milk. Final concentration of Fab
fragments was 0.l0 mg/ml.

bAverage of triplicate analyses.

87

interaction involving the Fc portion and various effector

functions.

Creaming in Simulated Milk Containing

 

Synthetic Milk Fat Globules

 

If the reaction is specific, the MFG and in particular
the milk fat globule membrane (MFGM) should play an active
role in the fat globule clustering process, as opposed to
a passive one in which the IgM merely precipitates onto its
surface. To test this hypothesis, synthetic fat globules
were prepared from butter oil and a stabilizing agent of
either K-casein, or an 05-, B-casein—rich fraction, or a
mixture of these two fractions. Synthetic fat globules and
washed MFG were added separately to heated skim milk. IgM
was added to one half of each sample and SMU to the other
half which served as controls. Results in Table I3 show
that all samples without IgM failed to form fat globule
clusters or a cream layer. This was expected with the MFG
and demonstrates that the synthetic fat globules do not
undergo non-specific interaction or excessive gravitation.
In the samples containing IgM, only the one with MFG formed
normal fat globule clusters and a cream layer. This supports
the concept that a specific interaction is involved in

which both IgM and MFG play an active role.

.00000000 00000—000 mo 0m000><0

.co00000w 000000 000000o0000 000 000000000 0 0:0 .0

.00 .0m.m 0o 0:00co0 00+ 00:00 0 o0 00000 0003 00: .0F0E 2000 0o 00000 m.0 o0 00000
0003 000000o0 0—00o—m 00» mo 0000 00o 0:0 02m 0o :o0000o0 020-200 0o 0000 00o 0000000

 

88

 

 

+ 00 0N.0 0_.0 00.0 00.0 00-0\0\00 +

+ 00 0N.0 00.0 00.0 00.0 00-0 +

+ 00 0N.0 00.0 00.0 00.0 00-0\00 +

0.00 0N.0 00.0 00.0 00.0 00: +
200 + 5000 00000:

+ 00 0N.0 00.0 00.0 00.0 00-0\0\00 +

+ 00 0N.0 00.0 00.0 00.0 00-0 +

+ 00 0N.0 00.0 00.0 00.0 00-0000 +

+ 00 0N.0 00.0 00.0 00.0 00: +
0000 000000

. 0 0N 0 0 0 N 0 0

00000 00000 0000000 000 00\_sv 0000003 00000 0000000

 

.Ao0zv 000:0oFm 000 0005 000000000 0000000co0 0003 00000000coo00 :0 00050000 .m— 0—000

89

Based on the ability of cows' milk euglobulins to
cluster fat globules from goats' milk and Holland-Friesan
and MRY cows' milk, Stadhouders and Hup (1970) concluded
that the euglobulin-fat globule complex was not specific.
Their conclusion fails to acknowledge the complexity,
heterogeneity, and microheterogeneity of the carbohydrates
on plasma membranes and hence milk fat globule membrane

(MFGM) and the cross-reactivity of antibody molecules.

The Effect of Soluble Milk Fat Globule

 

Membrane Proteins on Creaming

 

If an antigen-antibody reaction is involved, the agglu-
tination reaction should be inhibited by the addition of
soluble antigen - in this case soluble MFGM proteins.
Soluble MFGM proteins were isolated according to the pro-
cedure of Herald and Brunner (1957). Their addition to
raw milk did decrease cream volume and increase the time
for visible clusters to be formed (Table l4). To verify if
a specific effect was involved, ovalbumin was added to raw
milk and found to be without effect. The results further
support the notion that a specific interaction between IgM
and the MFGM is involved in the cluster/creaming mechanism.

In two-dimensional double diffusion immuno-analyses of
IgM versus soluble MFGM protein, precipitin lines failed to

develop. This was attributed to either improper titers or

90

.00000000 000000000 00 0m000><0

 

 

 

00.0 00.0 00.0 00.0 _E\0e o0 .cws=0_0>o +
00.0 00.0 00.0 00.0 _s\me 00 .200z-m +
00.0 00.0 00.0 00.0 05\0E 0 .200z-m +
00.0 00.0 00.0 00.0 Fe\me N .200z-m +
00.0 00.0 00.0 00.0 0005 200
g 00 g 0 g N c 0

0:05 0500 0000000 000E0m

A05 op\_ev 00E:_o> E0000

 

.000E mo

00050000 co 00000000 02002-00 00000505 0000000 000 0005 0000000 mo 000000 000 .00 00000

91

the formation of soluble complexes. In instances where
the antigen is bivalent or monovalent, precipitation may
not occur due to the lack of formation of large aggregates

or lattices.

The Effect of Milk Fat Globule Membrane Specific Anti-

 

sera on Creaming of Heated Milk

 

To further verify that antibodies specific for MFGM
antigens are involved, anti—serum against MFGM was prepared
in rabbits and added to heated milk to examine its ability
to restore creaming. Two different anti-serum preparations
were employed. As a control, anti-serum against navy bean
trypsin inhibitor (NBTI) was also prepared. The results in
Table l5 show that only the anti-MFGM hyperimmune anti-sera
restored normal creaming to heated milk. The limited clus—
tering observed with anti-NBTI anti—serum after 3l.5 min
may be attributed to the presence of antibodies normally
present in rabbit serum. These results support the belief
that MFG clustering is the result of a specific interaction
(antigen-antibody) between IgM and the MFGM.

In a study of the molecular mechanisms of milk secretion
in which probe molecules were infused into the lactating
gland of the goat via the teat canal, Patton _t _l. (l980)

employed goat MFGM specific anti-serum. They observed that

milk from the anti-serum-infused side of the udder showed

92

.N 00000000000

.0 00000000000

.00000000 000000000

50000-00:<0

50000-00:<

2

0o 0m000><0

 

 

 

m.m om.— mo.m om.¢ mm.~ waE 30m
m.~m 00.0 mfi.o O_.o o—.o —E\Hhmznwp:m FE OP.O +
0.0 —E\uzwmzuwpcm FE OF.O +
N.NF —E\Uzwmzlwwcm FE mo.o +
m.© —E\nzwmzlwucm FE OF.O +
m._~ m¢.~ O©.— 00.0 o—.OF —E\nzwuZuwp:m FE mo.o +
+ 00 om.o op.o OF.O .o—.o x—wE Uwpmwz
: «N c m n N 5 F
0:050 00500 0000000 005 00\_5V 005:_o> 50000 00050m

 

.0005 00000; 00 00050000 :0 0000-0000 0005500000; AHHmzv 000000000
0000000 :000 >>0cu00c0 0:0 Azw0zv 00000505 0000000 000 0005:0000 0o 000000 000 .m0 00000

 

93

extensive clustering of fat globules after storage for l2-24
h at 4 C. Milk from the non-infused side exhibited essen-
tially no clustering. This observation is consistent with
the suggestion of Jenness and Parkash (l97l) that goats'
milk creams poorly at low temperatures due to the lack of

agglutinating euglobulins.

The Effect of Simple Sugars on Creaming

 

Since IgM is involved in fat globule clustering, and
cows' milk generally creams throughout the lactation cycle,
one may expect natural antibodies - or antibodies associated
with histocompatability antigens or blood groups - to be
involved. The isoagglutinins, which are predominantly
IgM, occur with predictable regularity in all individuals
without any apparent overt antigenic stimulation. They
normally appear in serum 3 to 6 months after birth and
remain throughout the life cycle. Their titer may vary
throughout this period.

If natural antibodies are involved in the interaction,
the antigen would be carbohydrate in nature. To test this
possibility, hapten-inhibition studies, involving the
addition of simple sugars to milk, were performed. The
inhibition assay is essentially a competitive one in which
the inhibitor competes with the native antigen on the fat

globule surface for antibody molecules. The greater the

94

similarity between the added carbohydrate and the antigen,
the greater the degree of cross-reactivity, and the greater
the degree of creaming inhibition. When a number of simple
sugars were tested, only glucosamine, galactosamine, and
sialic acid were found to have an inhibitory effect

(Table 16). In some milk samples sialic acid was found to
completely inhibit creaming. The variability of response
to the addition of sialic acid is a ramification of antibody
heterogeneity. The complete and greater inhibition of
creaming with sialic acid than with glucosamine or with
galactosamine indicates sialic acid may be more closely
related to the immunodominant group(s). The concentration
of sugars required reflects the lack of identity with the
MFGM antigen, but the notion of cross-reactivity does
reflect the nature of the antigen.

That the antigen is carbohydrate in nature is supported
by the observation that heat treatment does not destroy
antigenic activity. Heated cream has been shown to cream
as well as non-heated cream when added to raw skim milk

(Dunkley and Sommer, l944).

Fat Globule Clustering Interpreted in Light

 

of Immunologic and Serologic Knowledge

 

It is interesting to speculate on the scenario drawn

by the above experiments in light of information Which has

95

.00000000 000000000 00 0m000><0

.000000 0000003 0003 05\m5 00 00 0000000000000 0 00 0:0500o0000000000uz 0:0 .0005
-000000000000000uz .000000 .00000x .00o00000m .00occ05 .0000000 .000000— 00 00000000 0000

 

 

 

0.00 00.0 00.0 00.0 00.0 _E\02 0.0 .0002000000000 +
0.00 00.0 0N.0 00.0 00.0 _5\05 0.0 .00050000000 +
0.00 00.0 00.0 00.0 00.0 FE\0E 0.0 .0000 000000 +
0.00 00.0 00.0 00.0 00.0 _5\05 0.0 .0000 000000 +
0.00 00.0 00.0 00.0 00.0 _5\05 0.0 .0000 000000 +
0.0 00.0 00.0 00.0 00.0 _5\05 0.0 .0000 000000 +
0.0 00.0 00.0 00.0 00.0 0005 200
0 «N 0 m g N s 0
00050 00500 00000—0 005 00\_50 005:_o> 50000 000500

 

.00005 00 00050000 :0 000000 000500 00 000000 000 .00 00000

96

been collected on the human and cattle immunologic and
serologic systems. Cold agglutinins or antibodies more
active (higher titer) in hemagglutination at lower tem-
peratures are normally found in all fresh blood serum (Race
and Sanger, l968). Clinical symptoms are generally not
present. Cold agglutinins are generally associated with
the A, H, Lea, P, I, Sp], M, and N antigens (Zmijewski and
Fletcher, l972). It is significant with respect to fat
globule clustering that A and B antigen-associated anti-
bodies have been detected in milk (Race and Sanger, l968).
The presence of antibodies associated with other antigens
can therefore be expected. This could explain the specifi-
city and temperature-dependence of the reaction between MFG
and IgM.

The following information is presented in support of
the conjecture that it is possible for the I—antigen and
its associated IgM antibodies to be involved in MFG clus—
tering. However in so doing, it should not be construed
that the I-antigen is involved or that it is the only
possible antigen involved. The I-antigen is widely distri-
buted and considered a "public" antigen. It generally
appears on the erythrocyte membrane in the first months
of life, replacing the I-antigen observed at birth. Anti-I
antibodies are of the IgM class and are essentially complete
natural antibodies possessing an optimum activity at 4 C.

After fixation on erythrocytes at 4 C they can be completely

97

eluted at 37 C. Because anti-I agglutinins occur in sera
of individuals who also have the I-antigen, they can be
considered autoantibodies.

I-active material (antigens) can be produced by the
stepwise degradation of A and B substances. Thus, rather
than being something separate, the I-antigens are internal
to the A, B, H, Lea, and Leb determinants, and the I-gene
exerts its influence prior to that of the ABH and Le genes
(Race and Sanger, l972; Goudemand and Delmas-Marsalet,
l975; Zmijewski and Fletcher, l972).

Because blood group substances or antigens are present
on nearly all tissues, it is expected that they are present
on the apical section of the mammary gland secretory cell
and hence on MFG. While possessing cell-type-specific
antigens (Ceriani _t _l., 1977), numerous common complex
carbohydrate species have been identified on the MFG
(Glockner gt_al., l976; Newman _t 31., l976; Newman and
Uhlenbruck, l977; Farrar and Harrison, 1978). Complex
carbohydrates corresponding to A, M, N, TF, and Tn antigens
have been identified. HLA-DR-like antigens have also been
found on the MFG (Niman _t _l., l979). Due to the extensive
microheterogeneity displayed by glycoproteins and glyco-
lipids, the presence of carbohydrate moieties similar or
identical with the I-antigen are possible.

Curtain (1969) found IgM cold agglutinins, mostly with

anti—I specificity, in high incidence and titer in sera

98

from sheep, cattle, kangaroos, wallabies, wombats, and
possums. Since all the cells tested from the various
species possessed some I-antigen, the IgM species could be
regarded as autoantibodies. Cattle erythrocytes had the
lowest concentration of I-antigens, but their sera had the
highest incidence and titers of cold agglutinins.

The presence of autoantibodies does not imply the
presence of an autoimmune disorder. Since an antigen only
induces an immune response when it is ”foreign to circula-
tion” or ”foreign to antibody forming cells", anatomic
location can prevent an antigen from being recognized as
self. Tissue extracts, e.g., brain, kidney, testis, and
crystalline lens of the eye, can induce an immune response
when injected into animals of the same species from which
they originated. This is especially true if the antigens
occur intracellularly. The immunologic response to auto-
logous, organ—specific antigens generally has no untoward
consequences for the immunized animals (Milgrom, l969;
Kabat, 1976).

Wiener gt_al, (l956) suggested that cold hemagglutinins
might be cross—reacting antibodies to antigenic determi-
nants derived from organisms and substances in the environ-
ment, but requiring cooling to express their specificity
towards the I-antigen. Curtain (l969) suggested that
possibly the I-antigen, rather than anti-I, is inactive at

37 C. It is therefore not recognized as an autoantigen and

 

99

immunological tolerance to it and related configurations
does not occur, enabling high titers of anti-I to be
reached. If this postulation is correct, anti—I cold
agglutinins are not autoantibodies, as the I-antigen would

not exist at physiological temperature.

The Effect of Enzymic Treatment of Milk Fat

 

Globules on Creaming

 

The above experiments indicate that the MFGM, and in
particular its carbohydrate moieties, play an active role
in fat globule clustering. One would therefore expect fat
globule clustering to be effected if carbohydrate moieties
are removed or modified enzymically. Isolation and frac-
tionation of glycopeptides released with proteases could
lead to the specific carbohydrate moiety or moieties
involved through inhibition assays.

Results obtained using pronase, trypsin, neuraminidase,
and mixed glycosidases are presented in Table l7. Treatment
with mixed glycosidases, neuraminidase, or a mixture of
these enzymes failed to significantly effect fat globule
clustering. Neuraminidase treatment increased cluster time
slightly while treatment with mixed glycosidases decreased
cluster time slightly. Protease treatment significantly
decreased cluster time. Trypsin treatment followed by mixed

glycosidase treatment reduced cluster time to 0.5 min.

 

lOO

.000000:0 000000000 00 0m000><0

.xm.m 00 0:00:00 000 00:00 0 00 00000 0003 00030000 000 00020

.0000:00 0:0 00 0000500 0000000
00000000 .0050N:0 0:0:000:00 00: 0000000 0002 0000000 0003 00000000 000 0005 00003
:0 0000500 .50000 00 000000 0005 5000 00000: 00 00000 00000000 000 0005 .0000 0000 :00

 

000000000000

 

 

0.00 00.0 00.0 00.0 00.0 00x00 + 0000000500002
0.0 00.0 00.0 00.0 00.0 0000000000_0 00x00 .5000000
0.0 00.0 00.0 00.0 00.0 000000000000: .5000000
0.00 00.0 00.0 00.0 00.0 000000000000 00x02
0.00 00.0 00.0 00.0 00.0 0000050000002
0.0 00.0 00.0 00.0 00.0 0005000
0.0 00.0 00.0 00.0 00.0 0000000
0.00 00.0 00.0 00.0 00.0 0000000
0 00 0 0 0 N 0 0
0000 00000 0000000 APE 0_\000 005000> 50000 0000000000 0000000 000 000:

 

.00005 00 0:050000 :0 00000000 000 0005 00 0:0500000 005»~:0 00 000000 000 .00 00000

 

 

101

The fai1ure of neuraminidase to increase c1uster time
to a greater extent than observed cou1d be re1ated to the
enzyme's ag1ycon specificity or sia1ic acid accessibiTity.
Tomich _t _l. (1976) demonstrated that gang1iosides of the
membrane are shie1ded from neuraminidase attack by membrane
proteins. This cou1d a1so be true for the mixed g1ycosi-
dases. The s1ight decrease in c1uster time may have been
due to exposure of specific carbohydrate moieties. Pro-
teases may have decreased c1uster time through their effect
on the primary (IgM binding to MFG) or secondary (MFG agg1u-
tination) phases of fat g1obu1e c1ustering (Discussed in
the section Interpretation of the Effect of Environmenta1
Factors on Creaming). Since a11 MFGM proteins, inc1uding
severa1 gTycoproteins, are not equa11y susceptib1e to
proteo1ysis (Mather and Keenan, 1975; Shimizu t a1.,

1979; Shimizu t 1., 1980), c1eavage of those which are

susceptibTe may have exposed specific carbohydrate moieties.
Gang1iosides, which are Tocated primari1y on the environmen-
ta1 face, may a1so have been exposed (Tomich _t _l., 1976).
C1eavage of sia1og1ycopeptides wou1d effect the secondary
phase of fat g1obu1e c1ustering by decreasing the zeta
potentia1 (Newman and Harrison, 1973; Harrison _3 al.,

1975). This wou1d faci1itate agg1utination by a11owing MFG

to approach one another in a c1oser re1ationship.

 

102

The Effect of Mi1k Fat G1obu1e Membrane

 

Gang1iosides on Creaming

 

As a11uded to in the above discussion, MFGM gang1io-
sides contain comp1ex carbohydrates which cou1d partici—
pate in fat g1obu1e c1ustering. Erythrocyte b100d group
antigens are primari1y associated with g1yc01ipids (Kabat,
1976). To test the possibi1ity that MFGM gangTiosides
participate in creaming, hapten inhibition assays were
performed using a gang1ioside fraction from MFGM (Tab1e 18).
The gang1iosides fai1ed to inf1uence creaming when added to
raw mi1k. This cou1d be accounted for if they 1acked the
specific carbohydrate moieties invo1ved in creaming or on
the basis of their 1ow effective concentration or inaccessi-

bi1ity to the reaction due to mice11arization.

Confirmation of the Mertens

 

and Samue1sson Effects

 

The experiments of Mertens (1932) and SamueTSsonl_t‘_l.
(1954) contributed significant1y to our understanding of
the homogenization-induced destruction of creaming of mi1k.
Mertens demonstrated that a size reduction of MFG was not
sufficient to exp1ain the phenomenon by showing that homo-
genized skim mi1k containing washed MFG a1so fai1ed to

cream. This experiment imp1icated a 1abi1e skim mi1k phase

component. Samue1sson gt al. showed that the agg1utinin in

 

 

 

 

 

 

 

 

 

 

 

   

.00000000 000000000 00 0m000><0

.000E0xo0000 000 00000000000000 00000000000

 

 

mw .0005 00 00000 0 00 00000 003 00003 000000000 00 00000000 00000000000 00 0000 000 0000000
0.0 00.0 00.0 0.00 0.00 :0 00.0 .000000000000 +
0.0 00.0 00.0 0.00 0.00 20 00.0 .000000000000 +
0.0 00.0 00.0 0.00 0.00 :0 00.0 .000000000000 +
0.0 00.0 00.0 0.00 0.00 0000 300
0 00 0 0 0 N 0 0

 

0:050 0500 0000000

0 050
0 APE o_\_EV 00E:_o> E0000 0 0 m

 

.0002 00 00020000 :0 020:2:00000 00000505 0000000 000 0002 00 000000 000 .00 00000

 

104

mi1k consisted of 2 components: heat- and homogenization-
1abi1e components. It was shown that both were necessary
for norma1 creaming. The homogenization-1abi1e component
has not been identified. Payens (1964) was unab1e to
detect any difference in the physica1 properties or c1us-
tering abi1ity of eug1obu1in isoTated from coTostrum prior
to and fo11owing homogenization. No proposed mechanism
exp1aining fat g1obu1e c1ustering has prescribed a ro1e to
the homogenization-1abi1e component.

Experimenta1 resu1ts verifying the Mertens and Samu-
e1sson effects are presented in Tab1e 19. As expected, the
samp1es containing on1y homogenized or heated skim mi1k
fai1ed to support creaming, whi1e a mixture of both creamed
to a 1imited extent on the addition of washed MFG. Creaming
was 1ess than that observed in the contro1 samp1e. However,
on1y one ha1f of the usua1 concentrations of the heat- and

homogenization-1abi1e components were present.

Examination of the Ro1e of K-Casein in

 

Homogenization-Induced Destruction of Creaming

 

Koops gt al. (1966) demonstrated that when eug1obu1in
was homogenized with K-casein its abi1ity to promote
creaming was destroyed. They hypothesized that a eug10bu-
1in-K-casein comp1ex was formed, or that K-casein was

adsorbed to the MFG in such a manner as to prevent

105

.00000000 000000000 00 0m000><0

.0m.m 00 0:00:00 000
00:00 0 00 00000 0002 002 .00x0E 0003 00005 5000 00N000moso; 000 00000; 00 005000> 000000

 

 

 

0.00 00.0 00.0 00.0 00.0 002 + 00000000000 + 00000:
+ 00 00.0 00.0 00.0 00.0 002 + 0000000000:
+ 00 00.0 00.0 00.0 00.0 002 + 00000:
0.0 00.0 00.0 00.0 00.0 0000 300
0 0m 0 0 0 N 0 0
A00EV 00500 0000000 005 op\—EV 000000> 50000 0000500

 

.00020 00000000 000 0000 000003 0000000000 0000 0000
00N000moso; 000 0005 E000 00000; 00 0000x0e 0 00 .0005 5000 00N000moeo; .0005
E000 00000; :0 00050000 - 0000000 0000000500 000 000000: 0:0 00 000005000000

.00 00000

   

106

eug1obu1in bridging. Since K-casein contains comp1ex
carbohydrate moieties (Whitney _t al., 1976) which are
simi1ar to those found on the MFGM (Newman and Uh1enbruck,
1977), this theory was attractive. It was thought that
homogenization may expose the carbohydrate moieties in such
a manner as to a11ow reaction with IgM. However, experi-
ments conducted with mode1 systems of heated who1e mi1k
combined with one of the fo11owing: (a) homogenized
K-casein, (b) K-casein, (c) homogenized IgM, (d) IgM,

(e) homogenized (K-casein + IgM), (f) K-casein + IgM,

(9) homogenized K-casein + IgM, or (h) K-casein + homoge-
nized IgM, fai1ed to support their observations. Substitu-
ting caseinomacropeptide or mice11ar casein for K-casein or
increasing homogenization pressure to 2500 psi fai1ed to
change the resu1ts obtained. NaTstra (1980) has shown that
homogenization of whey, either rennet or acid, inhibits the
abi1ity of the whey to support creaming and sheds doubt
upon an IgM-K-casein comp1ex being formed during homogeni-
zation. The author did not propose an a1ternative mechanism
of inactivation. A1though whey was not a very good creaming
medium (Discussed in the section Creaming in Mode1 Systems -
Simu1ated Whey Mode1 Systems), these resu1ts were reproduced

in this study (Tab1e 20).

 

107

.00000000 000000000 00 0m000><0

.00.0 00 0:00:00 000 00000 0 00 00000 0003 0020

 

 

 

+ om 0N.0 00.0 00.0 00.0 00: + 0003 000000 00N0c0mosoz
+ 00 00.0 0N.0 00.0 00.0 002 + 0003 000000
+ 00 00.0 00.0 00.0 00.0 002 + 0003 0000 00N00000201
+ om 00.0 00.0 00.0 00.0 002 + 00:3 000<
0 0m _ s m 0 N g 0
Ac0ev 00200 0000000 005 o_\_svn0E:—o> E0000 0000500

 

.Aw0zv 00000000 000 0002 000003
0000000000 00003 000000 000 0000 00N0cmmoeocnco: 0:0 00N0c0moEo; :0 00050000 .om 00000

 

108

Examination of the Participation of Skim Mi1k

 

Membrane in Creaming

 

Because the reaction between IgM and the MFG was shown
to be specific, IgM cou1d react with carbohydrate-containing
species other than MFG in mi1k. Two possib1e components are
mi1k o1igosaccharides — the soTub1e comp1ex carbohydrates
found in mi1k - and skim mi1k membrane (SMM) - much of
which has an origin simi1ar to the MFGM (P1antz et_al,,
1973; Kitchen, 1974; Patton and Keenan, 1975). Due to the
10w 1eve1 of mi1k o1igosaccharides in cows' mi1k (B1anc,
1979), their possibie participation was not examined. The

possib1e participation of SMM was examined.

Effect on Creaming of Homogenized Skim Mi1k Containing

 

Washed Mi1k Fat G1obu1es

 

The addition of SMM, isoiated by u1tracentrifugation
or sa1t fractionation and ge1 fi1tration, to homogenized
skim mi1k containing washed MFG produced the resu1ts repor-
ted in Tab1e 21. The resu1ts c1ear1y show that whi1e IgM
fai1ed to restore creaming, SMM did restore creaming to
homogenized skim mi1k containing washed MFG. A very 10w
c1uster time, very Targe c1usters, and a deep cream 1ayer
were obtained. The resu1ts indicate that SMM is the
homogenization-1abi1e component. The different c1uster
times for the two SMM preparations may be due to (a)

differences in quantity of SMM added to the mi1ks (due to

 

109

000 000 00000000000 003 00m

000 00 0 .0000000m 00m-000 00 m_.¥_00 0000 00 0.0

.0000000000 F00 000 0000000000000 000m >0 00000000 22m0

.0000000000000000000 >0 00000000 zzmu
.m0mxp000 0p0u_p0:0 we 0m000><0

.00000000 0003 zzm 000
0: 000m03 FE m.N 000 .00000000

.00000000 000 00:00 00.00 0
F 000x00 00 00000000 0003 m0_0s0m0

 

 

 

 

P.m_ mo.F om.F 0.00 0.00 022m +
+ 00 00.0 00.0 00.0 o_.o 002 + 0000 00N0=000000
m.~ mo.0 mm.m om.m 00.0 022m +
+ 00 00.0 00.0 00.0 00.0 _0\00 00.0 .000 +
+ 00 00.0 00.0 00.0 . 00.0 000 + 0000 00N_0000000
0 0m 0 m 0 N 0 F
A005 0500 00pm0~u AFE o_\_EV 00E:_o> E0000 00F0E0m
.Ao0zv m0rsnoFm p0» 0000 000m03 0000000000
JF0E wam 00N000moEO0 00 00000000 00 000000pFPw F00 000 0000000000000 000m
00 0000000000000000pF: 00 00000000 Azzmv 00000502 0000 500m mo 000000 000 ._N 00000

 

110

approximations and 1osses in preparation), (b) differences

in IgM content of the mi1ks empioyed, or (c) the re1ative

distribution of components isoiated by the two procedures.

Effect on Creaming of Heated Skim Mi1k Containing Washed
Mi1k Fat G10bu1es

 

Whi1e IgM did restore creaming to heated mi1k, SMM
fai1ed to do so (Tab1e 22). This further demonstrates that
SMM is distinct in function from the heat-1abi1e IgM

component.

Effect on Creaming of Gravity-Separated Skim Mi1k Containing
Washed Mi1k Fat Giobu1es

 

Based on the fai1ure to support creaming despite the
presence of active IgM, it was conc1uded that the fai1ure
of gravity-separated skim mi1k to support creaming was due
to the absence of another 1imiting component. To determine
if the 1imiting component is SMM, it was added to gravity-
separated skim mi1k prior to creaming experiments. The
resu1ts in Tab1e 23 show that the addition of SMM did
restore norma1 creaming to the gravity-separated skim mi1k.
As expected, IgM fai1ed to do so. These resu1ts further
indicate that SMM is the 1imiting component in determining
how many times a given skim mi1k wi11 cream and probab1y

a1so in the amount of IgM incorporated into the cream 1ayer.

 

1
1
1|

.m0m>_000 00000_000 0o 00000><0

.00000000 0003 220
000 200 000 00000000000 m0: 020 .Ap00p0ou 000 P0000 00.m0 002 000m03 _E m.N 000 .0000
-0_00 :20 _E 0 .0owp:_0m 020-200 PE 0 .0FwE wam FE 0.00 000wa 00 00000000 0003 m0P0E0m0

 

 

 

+ 00 om.o o—.o 00.0 o—.o 22m +
0.00 00.0 00.0 00.0 00.0 00000 00.0 .000 +
+ co mN.o 00.0 o—.o 00.0 00: + E00m 00pmm:
0 0m 0 m 0 N 0 —
A00Ev 00E00 00000—0 APE oF\FEv 00Espo> E0000 00F0E0m

 

.00020 m0_:aopm #00 000E 000m03 0000000000 000E E000 000000 mo 000E000u
0o 0000000000000000pF0 00 00000000 02200 00000E0E 0FwE wam 00 000000 000 .mm 0_00E

 

112

.0000000000 ~00 000 0000000000000 “—00 00 00000000 22m0

.0000000000000000p_: 00 00000000 2200
.00000000 00000P000 mo 00000><0

.00000000 0003 22m 000
:00 000 00000000000 003 :20 .00000000 000 _0000 0m.mv 002 000003 —E 0.0 000 .00000000
:20 PE 0 .00000000 020-200 PE 0 .xPWE E000 FE m.m_ mcwva 00 00000000 0003 00—0E0m0

 

 

 

N.© mm.— om.~ 00.0— 00.0_ 022m +
0.0m 00.0 om.o 00.0 o—.o 002 + wam pr>00o
F.m 00.0 ow._ mw.— om.— ozzm +
0.0m 00.0 00.0 00.0 00.0 0E\00 00.0 .000 +
+ co m©.o om.o 0N.0 00.0 002 + wam >pw>00w
0 0m 0 m 0 N 0 F
A00Ev 00Ewk 00000—0 APE o_\_EV 00E:_o> E0000 00P0E0m

 

.00000 00000000 000 0000 000003 0000000000
000E E000 000000000I>pw>00m 00 000E0000 00 0000000000 ~00 000 0000000000000
0000 00 0000000000000000000 00 0000Fomw Azzmv 00000E0E 0FPE E000 mo 000000 00» .mm 0F00E

 

113

Creaming in Mixtures of Gravity-Separated, Homogenized, and
Heated Skim Mi1ks Containing Washed Mi1k Fat G1obu1es

To further examine or verify the presence or absence
of IgM and SMM in various skim mi1k preparations, se1ected
skim mi1ks were mixed and examined for creaming abi1ity
(Tab1e 24). As previous1y shown, gravity skim - devoid of
or 1acking functiona1 SMM, but sti11 containing active IgM -
fai1ed to cream norma11y upon the addition of washed MFG.
A mixture of gravity-separated skim mi1k and homogenized
skim mi1k - both of which are devoid of or 1acking function-
a1 SMM, but containing active IgM - a1so fai1ed to cream.
When gravity-separated skim mi1k and heated skim mi1k -
which is devoid of active IgM - are mixed, significant1y
improved resu1ts were obtained. Creaming was not returned
to norma1 1eve1s, but on1y one ha1f of the usua1 concentra-
tions of the heat- and homogenization-1abi1e components
were present. These resu1ts support the contention that
gravity-separated and homogenized skim mi1ks fai1ed to
support creaming due to the absence of the homogenization-
1abi1e SMM, whereas heated skim mi1k fai1ed to support

creaming due to the absence of the heat-1abi1e IgM.

Effect on Creaming of Raw Mi1k

 

When SMM was added to raw mi1k the resu1ts presented
in Tab1e 24 were obtained. The cream 1ayer depth was in-

creased and c1uster time decreased. This observation

 

.00000000 00000—000 00 0m000><0

.00x0E 0003 0000E E000 00N000moE00 00 000000
000 >00>00m mo 00E000> 00000 .0m.m mo 0000000 000 00000 0 00 00000 0003 002 0000030

114

 

 

+ 00 00.0 00.0 00.0 00.0 E000 00N000moEo: +

0.0m 00.0 00.0 00.0 00.0 E000 00000: +

+ 00 00.0 00.0 00.0 00.0 00: + E000 >00>000
0 0m 0 m 0 m 0 0

 

A00Ev 0E00 0000000 0000E0m

00E o_\0EV 00E000> E0000

 

.00020 00000000 000 000E 000003 0000000000
0000E E000 000000 000 .00N00000E00 .000000000nx00>000 00 000:0x0E 00 000E0000 .0N 00000

 

 

115

.00000000 000000000 00 0m000><0

.0000E00 0000 00 00000000 00000 0E 000 00000000
00 0000E00 00000 000 0000000 000 00 00000 003 000000000000000000: >0 00000000 00030

 

 

 

N.¢ Om.m Ow.m 0.0_ 0.00 FE mN\FE m0 .sz +
¢.m m—.w om.m o.o_ o.o~ FE mm\FE OF 022m +
w.© 00.0 oo.m 0.00 0.00 FE mm\~E m .sz +
m.© OO.N mo.© om.n om.m waE 30m
c 0N 5 m c N s F
A00EV 00E0E 0000000 00E o_\_Ev 00E000> E0000 00—0E0m

 

.0—0E 0o 000E0000
0o 0000000000000000000 00 00000000 02200 00000E0E 000E E000 00 000000 000 .00 00000

 

 

 

116

further demonstrates that SMM does p1ay an active ro1e in

the fat g1obu1e c1ustering process.

Effect on Creaming of Co1d Mi1k

 

Co1d-agitated mi1k fai1$ to re-cream as exhaustive1y
as mi1k pre-warmed prior to p1acement at reduced tempera-
tures (Dunk1ey and Sommer, 1944). Mu1der and Waistra (1974)
suggested that co1d skim fai1s to support creaming due to
the formation of a few 1arge IgM cryoaggregates with
reduced abi1ity to f10ccu1ate fat gTobuTes. An ana1ogous
situation can be postu1ated for c01d mi1k which has previ—
ous1y creamed prior to agitation. However, since IgM has
been shown to function as a cryoaggTutinin as opposed to a
cryog1obu1in, this phenomenon was re—examined. IgM and SMM
were added to homogeneous con-agitated mi1k which was then
a110wed to re—cream (Tab1e 26). The samp1e to which SMM
was added creamed significant1y better than the con mi1k
or c01d mi1k suppTemented with IgM. The pre-warmed mi1k
creamed as expected. A possib1e reason for the poor re-
creaming of con mi1k and its improvement upon the addition
of SMM wi11 be discussed in the section, A Mode1 Represen-

tative of Mi1k Fat G1obu1e C1ustering.

Iso1ation of Skim Mi1k Membrane from Warm- and Coid-

 

Separated Skim Mi1ks

 

Based on the above experiments, co1d-separated skim

mi1k shou1d contain 1ess SMM than warm-separated skim mi1k.

117

.000>0000 00 00000 00E cm 000 0 mm 00 00E003 003 000E 0000 00 0000000 0<0
.000E0000 0000
0000000000 000000 003 000E E000 000 .000—E0000 0000000 000 00>00 E0000 000000 >0000000

.000E 0000 00 0E 00 00 00000
0000 003 0 0m 00 00000000 220 00 00000000 020-200 000000 00 0000000000E 030 .0000000000
N 0003000 000000000000 >0 00x0E 0000 000 0000000>0 0 0 00 E0000 00 0030000 003 000E 3000

.000>_000 000000000.00 0m000><0

 

 

 

0.0 00.0 00.0 00.0 0.00 00000 0003
0.0 00.0 00.0 00.0 00.0 000 +
0.00 00.0 00.0 00.0 00.0 000\00 00.0 .000 +
0.00 00.0 00.0 00.0 00.0 0.00000 0000
0 00 0 0 0 N 0 0
A00EV 00E00 0000000 00E 00\0Ev 00E000> E0000 000E00

 

.000E 0000 00 000E0000
00 0000000000000000000 >0 00000000 02200 00000E0E 000E E000 00 000000 000 .0N 00000

 

 

118

Re1ative quantities of SMM can be determined from the area
of the void voTume peaks in the Sepharose 4B chromatograms
of warm and con-separated skim mi1k (5,000)<g for 20 min)
as presented in Figure 10. The co1d-separated skim mi1k
contained 1ess SMM than the warm-separated skim mi1k. An
area ratio of 0.62 was obtained. The fact that the SMM
remaining in the co1d-separated skim mi1k wou1d not support
creaming indicates that not a11 SMM is capab1e of partici-
pating or is required in creaming. The cor011ary experi-
ment was performed to ascertain if the SMM content returned
to the origina1 1eve1 when co1d mi1k is warmed prior to
separation. An area ratio, re1ative to warm-separated skim
mi1k,0of 0.80 was obtained. Mechanica1 entrapment or
incomp1ete e1ution from the MFG may have resu1ted in the
1ower than expected va1ue. However, the trends support

previous1y presented resu1ts.

Iso1ation of Skim Mi1k Membrane from Non-Homogenized and

Homogenized Skim Mi1ks and Mode1 Systems

 

To acquire an insight into the inactivation of SMM
by homogenization, SMM was isoTated by ge1 fi1tration from
homogenized (2500 psi) and non-homogenized skim mi1ks and
mode1 systems - consisting of who1e casein and SMM iso1ated
from skim mi1k using ge1 fi1tration. A void vo1ume peak
area ratio (homogenized/non-homogenized) of 3.7 was

obtained for the skim mi1k, whereas for the mode1 system a

 

119

Figure 10. Ge1 fi1tration chromatograms of (A) warm-
separated skim mi1k (2.67 m1 samp1e) and (B)
co1d—separated skim mi1k (2.60 m1 samp1e). A
co1umn of Sepharose 48 (2.6 cm x 35 cm) was
emp1oyed.

120

2.00

1.00

 

 

 

 

 

 

am

 

 

2.00

1.00

 

 

 

 

 

 

 

20 40 60 80
FRACTION

Figure10

 

 

 

121

va1ue of 1.2 was obtained. The increased turbidity of the
skim mi1k void vo1ume fraction significant1y increased
absorbance at 280 nm. The non-homogenized skim mi1k void
vo1ume fraction contained 86% as much protein as the
homogenized samp1e. The non-homogenized SMM/casein samp1e
void vo1ume fraction contained 91% as much protein as the
homogenized samp1e.

These few resu1ts do not a11ow a mechanism of SMM
inactivation to be proposed. However, they do indicate that
the materia1 isoTated as SMM does change upon homogeniza-
tion. One can specu1ate that SMM is inactivated by (a)
emu15ification of 1ipoida1 membrane partic1es by proteins
in mi1k or (b) by a reduction in size of membrane partic1es.
In the first case carbohydrate reaction sites may be
shie1ded by adsorbed proteins, whi1e in the second size
reduction of partic1es may 1ead to a decrease in the number
of reactive sites/partic1e and hence a reduced abi1ity to
act as bridging-agents and possibiy even to acting as
inhibitors of creaming (Discussed in the section, A Mode1

Representative of Mi1k Fat G1obu1e C1ustering).

Effect of Rep1acing Skim Mi1k Membrane with Mi1k Fat

 

G1obu1e Membrane

 

Because SMM and MFGM have origins in common (P1antz

t 1., 1973; Kitchen, 1974; Patton and Keenan, 1975) it was

suspected that MFGM obtained by churning washed MFG may

122

function in restoring creaming capacity to gravity-separated
and homogenized skim mi1ks in the same fashion as SMM.
However, when added at various 1eve1$ to who1e mi1k,
homogenized skim mi1k, or gravity-separated skim mi1k it

had no effect. This behavior may have been a resu1t of
differences in reaction site (specific carbohydrate moie-
ties) density or 1ack of physica1 avai1abi1ity of reactive
sites due to different states of aggregation of vesicu1a—

tion.

Creaming in Mode1 Systems

 

Simu1ated Whey Mode1 Systems

 

The resu1ts in Tab1e 27 demonstrate that acid, rennet,
and centrifuge whey poor1y support creaming. The addition
of SMM and to a 1esser extent IgM improved their creaming
capacity. In generaT, centrifuge whey was a better medium
than rennet whey which was better than acid whey. When
subjected to ge1 fi1tration, void voTume peak area ratios
(whey/skim mi1k) were: 0.0 for centrifuge whey, 0.23 for
rennet whey, and 0.11 for acid whey. These ratios indicate
that the wheys 1ack functiona1 SMM as was anticipated for
centrifuge whey. The 10w pH emp10yed to precipitate casein
probab1y resu1ts in SMM precipitation in acid whey (Brunner,
1974) and physica1 entrapment of the SMM in the casein c10t

may account for its decrease in rennet whey. Due to the

123

 

.000>0000 000000000 00 00000><0

.00000000 00» 00000 00.mv 00020 00000000 00% 000E 000003 0E m.N 000 .00000000
02:00 00000E0E 000E E000 0E 0 .00000000000000 00000 _E\0E 00.00 00000000 020-200 0E m.N
.>003 0E 0.0N 000x0E >0 00000000 0003 0000E0m

.0.0 :0 00 00000000 0003 0000E00 >0030

 

 

 

 

0.00 00.0 00.0 00.0 00.0 000 + 000 +
0.00 00.0 00.0 00.0 0.00 000 +
0.00 00.0 00.0 00.0 00.0 000 +
+ 00 00.0 00.0 00.0 00.0 000 + 0003 0000000000
0.00 00.0 00.0 00.0 00.0 000 + 000 +
0.00 00.0 00.0 00.0 00.0 000 +
0.00 00.0 00.0 00.0 00.0 000 +
+ 00 00.0 00.0 00.0 00.0 000 + 0000 000000
0.00 00.0 00.0 00.0 00.0 000 + 000 +
0.00 00.0 00.0 00.0 00.0 000 +
0.00 00.0 00.0 00.0 00.0 000 +
+ 00 00.0 00.0 00.0 00.0 000 + >003 0000
0 0m 0 0 0 N 0 0
00EV 00E0H 0000000 APE o_\0Ev 00E000> E0000 0000E0m
.0E000>0 0000E >003 000000E00 00 000E0000 .NN 00000

 

124

10w SMM contents, MFG c1ustering in whey is primari1y IgM
mediated (Discussed in the section Creaming in Mode1
Systems - Simu1ated Mi1k U1trafi1trate Mode1 Systems).

The ranking of their capacity to support creaming is anti-
para11e1 to their ionic sa1t content. Since the addition
of saTt to mi1k has been shown to decrease creaming
(Dunk1ey and Sommer, 1944), this is probab1y a contributing
factor. Dia1ysis of the various wheys versus SMU may

equaTize their capacity to support creaming.

Simu1ated Skim Mi1k Mode1 Systems

 

The importance of SMM in creaming is further demon-
strated by the resu1ts of Tab1e 28. Simu1ated mi1k
prepared from centrifuge whey, casein mice11es, and washed
MFG creamed better (characteristics more 1ike the raw mi1k
contro1) in the presence of than in the absence of SMM.
The samp1e to which SMM was not added contained some SMM
attributab1e to that co-pe11eted with the casein mice11es.
However, the resu1ts indicate the importance of SMM in the

”native" mi1k system with respect to fat gTobu1e c1ustering.

Simu1ated Mi1k UTtrafi1trate Mode1 Systems

 

In order to ascertain the concerted and individuaT
contributions of SMM and IgM to creaming, washed MFG were
mixed with SMM, IgM, or a mixture of these two components
in SMU and creaming capacity eva1uated (Tab1e 29). The

resu1ts indicate that on1y the samp1e containing SMM and

 

125

.220 0000000 00 0000000000 00000 00000000000000 00000 00x0E 0003 00000000 0000 000000 0000

.0m.m 00 0000000 000 00000 0 00 00000 0003
00020 00000000 000 000E 000003 .00>0_ 00:00 02:00 00000E0E 000E E000 000 00 00>0E00
00000 >003 0000000000 00 00000 0003 000000000000000000: >0 00000000 0000000E 0000000

.000>0000 000000000 00 0m000><0

 

0000 + 000 + 000000 +

 

 

0.0 00.0 00.0 00.0 00.0 >002 0000000000
0002 + 000000 +
0.0 00.0 0N.0 00.0 00.0 >003 0000000000
0.0 00.0 00.0 00.0 00.0 000E 300
0 0N 0 m 0 N 0.0
000EV 00E00 0000000 APE o—\0EV 00E000> E0000 000E0m

 

.E000>0 0000E 000E E000 000000E00 0 00 000E0000 .mN 00000

 

 

 

   

126

.00000000 000E0 000 0000E0000 0000000 >00>0

.000>0000 000000000 00 00000><0

.00000000 000 00000 00.00 00000 00000000
000 000E 000003 _E m.N 000 .00000000 02200 00000E0E 000E E000 PE 0 .00000000000000

 

00000 _E\0E 00.00 00000000 020-200 _E 0 .020 _E 0.00 000x0E >0 00000000 0003 0000E0m0

 

 

0.0 00.0 00.0 00.0 00.0 000 + 000 +

0.00 00.0 00.0 00.0 0N.0 0220 +

0.00 00.0 00.0 00.0 00.0 0000 +

+ 00 om.o 00.0 00.0 00.0 00: + 020
A00Ev 00E00 0000000 0 0N APE “FWFEV 00E000M M0000 0 0 0000E0m

 

.E000>0 0000E 000000000000: 000E 000000E00 0 00 000E0000 .mN 00000

   

127

IgM disp1ayed norma1 creaming. The samp1e with SMM creamed
to a 1imited extent - probab1y due to the sma11 amount of
IgM present in the whey carrier. The samp1e of IgM creamed
to a 1imited extent - indicating IgM has the abiTity to
c1uster MFG to a 1imited extent in the absence of SMM.

The resu1ts of this experiment are simi1ar to those
obtained by Gammack and Gupta (1970). The authors noted

that 1ipoprotein partic1es iso1ated from mi1k augmented

 

creaming. A1though the partic1es were not recognized as

SMM, they were probab1y SMM or some fraction thereof.

Interaction of Mi1k Fat GIobu1es

with Skim Mi1k Membrane

 

To determine if SMM interacts with MFG without promo-
ting c1ustering, heated who1e mi1k was washed 4 times with
washed MFG at 4 C. After quiescent storage at 4 C for 6 h,
the heated mi1k was centrifuged (1000><g for 20 min) and
the MFG-containing 1ayer removed. Fresh washed MFG were
added to the skim 1ayer and the process repeated three
times. Heated mi1k served as a contro1. Resu1ts are pre-
sented in Tab1e 30. The addition of IgM a10ne produced
sufficient1y simi1ar resu1ts in the heated mi1k and MFG-
washed heated mi1k to conc1ude that SMM did not interact
with MFG. -

 

128

00.00 003 000003 00
_0\00 00.00 00000000

.000>0000 000000000 00 0m000><

0
.00000000 000 00000

0E m.m 000 .>003 0000000000 00 :20 00 0E m .00000000000000 00000
020-200 00 0E m .000E 00 0E 0.0N 000x0E >0 00000000 0003 0000E0m0

 

 

0.0 00.0 00.0 00.0 00.0 220 + 200 +

0.0 00.0 00.0 00.0 0N.0 :00 +

+ 00 0N.0 00.0 00.0 00.0 000E 000003-002 00000:

0.0 00.0 00.0 00.0 00.0 030 + 300 +

0.0 00.0 00.0 00.0 00.0 :00 +

+ 00 0N.0 00.0 00.0 00.0 000E 00000:
0 0N 0 m 0 N 0 0

 

A00EV00E00 0000000

00E 00\—Ev00E000> E0000

0000E0m

 

.000E 000000 0000031A00zv 0000000 000 000E
000 000E 000000 00 000E0000 00 02200 00000E0E 000E E000 000 200 00 000000 000 .00 00000

 

 

129

Interaction of Immunog10bu1in M with

 

Skim Mi1k Membrane

 

The precipitation of the heat- and homogenization-
1abi1e components from whey stored at 4 C (Samue1$son _t al.,
1954) is taken as evidence of an IgM-SMM interaction.

Perhaps this behavior can be considered as a typica1
antigen-antibody aggTutination or precipitation reaction.
The absence of a precipitate in the whey of poor-creaming
mi1k (Dunk1ey and Sommer, 1944; Bottazzi and Zacconi, 1980)

cou1d be attributed to an antigen or antibody (pro-zone

effect) excess.

Interpretation of the Effect of Environmenta1

 

Factors on Creaming

 

Dunk1ey and Sommer (1944) demonstrated that sa1ts are
required for MFG c1ustering, but excessive 1eve1s resu1t
in inhibition of c1uster formation. The ionic strength of
mi1k is higher than optimum as evidenced by the fact that
di1ution (up to 50%) favors the formation of a deep cream
1ayer and rapid formation of MFG c1usters. Po1yva1ent
cations added to raw mi1k decreased creaming to a greater
extent than monova1ent cations. Sodium citrate and disodium
phosphate had very TittTe inf1uence on creaming. Creaming
capacity of mi1k was reduced as the pH was varied from the

origina1 va1ue of 6.6. These observations were supported

 

130

by the resu1ts of Rhee (1969).

It was expected that sodium citrate and disodium phos-
phate might decrease the creaming properties by increasing
the zeta potentia1 of the fat gIobuTes and that CaC12,

A1C13 and FeC13 might improve creaming by decreasing the
potentia1. An improvement in creaming was a1so expected as
the pH was 1owered, due to a decrease in zeta potentia1.
Based on e1ectrophoretic mobi1ity measurements of fat gIo-
bu1es and the above resu1ts, the author conc1uded that the
charge on the MFG is not an important factor in the creaming
of mi1k.

In examining the effect of environmenta1 factors, i.e.,
ionic strength, pH, and die1ectric constant, on fat gIobu1e
c1ustering it is convenient to visua1ize the process as
occurring in two steps: the first, corresponding to the
uptake of IgM by MFG or SMM and, the second, the aggregation
of the MFG by the formation of inter-MFG (or SMM) protein
bonds (P011ack _t _l., 1964; P011ack, 1965; Zmijewski and
F1etcher, 1972). The two stages probab1y do not occur in
isoIated steps once the first stage has progressed suffi-
cient1y for the second to begin. A fat gIobu1e c1uster
wi11 resu1t on1y if the MFG can approach each other c10se1y
enough for the intervening distance to be spanned by the
IgM or IgM-SMM comp1ex. This distance wi11 vary with the
effective 1ength of the IgM moTecuIe or the IgM-SMM comp1ex

and the position of the antigen receptor site on the MFG.

 

 

 

 

 

 

131

The increase in the effective 1ength mediated by SMM may
contribute to its augmentation of MFG c1ustering.

Whether or not aggregation occurs is dependent on the
sum of the individua1 forces which drive the MFG together
and those which drive them apart. In the absence of gravi-
tationa1 and centrifuga1 forces, the major aggregating
force is through interfacia1 tension. With every surface
there is associated a definite amount of energy (y ergs/
cm2). When c1ustering occurs, two surfaces each of area
A cm2 are 10st and the amount of energy contained by these
two surfaces (ZA'yergs) becomes free. The MFG surface ten—
sion acts to produce c1ustering by reducing the free energy
to a minimum (equi1ibrium). If the surface is extensive1y
hydrated, the energy 1ost is not significant, but there is
no reason to assume that the MFG surface is extensiveTy hy-
drated or that the surface is inf1uenced to a marked extent
by the binding of IgM (Dunk1ey and Sommer, 1944; Payens,
1964; Payens _t al., 1965). The e1ectrostatic force of
repu1sion, which is dependent on the zeta potentia1, is the
determining factor in maintaining two MFG a given distance
apart. The zeta potentia1 is proportiona1 to the surface
charge density and the thickness of the doub1e 1ayer. It is
therefore inf1uenced by changes in the suspending medium,
i.e., ionic activity and die1ectric constant, which may
a1ter the doub1e 1ayer without a concomitant change in the

surface charge. An increase in the ionic strength or

 

 

 

 

   

132

die1ectric constant brings about a decrease in the doub1e
1ayer thickness and zeta potentia1. The term "critica1 zeta
potentia1" has been used to define the vo1tage above which
agg1utination cannot occur. For human erythrocytes, the
critica1 zeta potentia1 is about 18-20 mV for sa1ine agg1u-
tinins and 8-9 mV for incomp1ete Rh antibodies. For bovine,
dog, and mouse erythrocytes which are se1dom aggTutinated,
the critica1 zeta potentia1 wou1d be 1ower.

Factors inf1uencing the first stage of c1ustering or
the uptake of IgM by MFG or SMM are the ionic strength,
die1ectric constant, pH, and temperature of the suspending
medium. There is genera11y an optimum sa1t concentration
for antibody binding above which binding is decreased. As
pH is varied from the optimum va1ue of 6.7, antibody binding
is reduced. The optimum die1ectric constant depends on the
ionic strength of the medium but a decreased va1ue wi11
reduce antibody binding. Antibody binding is favored by
reduced temperatures.

From the above discussion, it is obvious that the inf1u-
ence of changes in pH and ionic strength are such that the
first and second stages of MFG c1ustering are inverse1y
affected. A decrease in pH or an increase in ionic strength
is detrimenta1 to the first stage but favors the second.

The effect of factors on the two stages of fat g1obu1e
c1ustering and cream 1ayer formation are difficuTt to

separate. Optimum fat g1obu1e c1ustering and creaming

 

133

occurs at conditions sub-Optima1 for the two individua1
stages which comprise the whoTe. This 1ed Dunk1ey and
Sommer (1944) to conc1ude that the charge on the MFG is

not important in fat g1obu1e c1ustering. The decreased
creaming capacity of mi1k on sa1t addition and decreased pH
are probab1y due to the reduced IgM-MFG interaction, whi1e
the requirement of sa1ts for c1ustering is probab1y asso-
ciated with the zeta potentia1. A decreased zeta potentia1
on remova1 of sia1og1ycopeptides probab1y contributed to
the decreased c1ustering times observed with protease-

treated MFG.

A Mode1 Representative of Fat

 

G1obu1e C1ustering

 

The most recent and genera11y accepted theory of fat
g1obu1e c1ustering (Brunner, 1974; Mu1der and Wa1stra,
1974) attributes agg1utination to the precipitation of
cryogIobuTins (IgM) onto the fat g1obu1es, 1eading to
c1ustering and rapid creaming of the 1arge c1usters
(Payens, 1968; Payens and Both, 1970). The mode1 can be.
pictoria11y presented as in Figure 11. Important features
are:(a) the precipitation of aggregated IgM onto MFG,and
(b) aggregates of IgM which precipitate in co1d whey. The
mechanism does account for the c1ustering of fat g1obu1es.

The temperature prerequisite is attributed to the nature of

 

 

 

Figure 11.

134

 

A mode1 representing an interpretation of mi1k
fat g1obu1e c1ustering as presented in the
contemporary Titerature. Circ1es and stars

represent mi1k fat g1obu1es and skim mi1k
membrane, respective1y.

 

 

 

 

 

 

 

 

135

the cryog1obu1ins. The effect of heat wou1d be to denature
the proteins which wou1d no 1onger function as cryog1obu1ins.
The proposed mechanism does not exp1ain why aggregated IgM
associates with MFG, as opposed to casein mice11es. It
a1so can not account for the Mertens and Samue1$son effects.
One of the major prob1ems appears to be that it fai1s to
prescribe a ro1e to the homogenization-1abi1e component.

A mode1 representative of fat gIobuTe c1ustering based

on the experiments and discourse of this dissertation is

 

presented in Figure 12. Significant features of the mode1
are: (a) three components are invoTved - IgM, SMM, and

MFG interacting through specific carbohydate moieties,

(b) IgM interacts with MFG and SMM, not SMM with MFG,

(c) IgM can c1uster MFG to a 1imited extent, (d) SMM acts
as a foca1 point or cross-Tinking agent which reduces the
distance which MFG must approach to a11ow cu1stering,

(e) IgM is free in so1ution, not cryoaggregated, and (f)
IgM can react with SMM free in so1ution - yiering the
precipitate in con whey containing the heat- and homogeni-

zation-1abi1e components.

As an app1ication of the proposed mode1, Figure 13
contains a pictoria1 presentation of events proposed to
occur in co1d-agitated mi1k and coId—agitated mi1k supp1e-
mented with SMM. In the absence of added SMM, the free

IgM in soTution competes with IgM attached to MFG and

 

 

136

Figure 12. A proposed mode1 of mi1k fat gIobu1e c1ustering
based on the experiments and discourse of this
dissertation. Circ1es, stars, and spira1
segments represent mi1k fat g1obu1es, IgM, and
skim mi1k membrane, respective1y.

 

Figure 13.

137

A pictoria1 presentation of events occurring in
co1d-agitated mi1k in the absence and presence
of supp1ementa1 skim mi1k membrane. Circ1es,
stars, and spira1 segments represent mi1k fat

g1obu1es, IgM, and skim mi1k membrane, respec—
tive1y.

 

 

139

associated SMM for the free carbohydrate moieties on SMM
and MFG. This resu1ts in fewer "bridges" and hence reduced
c1ustering and poorer creaming. In the presence of added
SMM, a situation simi1ar to that in warm mi1k is obtained

and numerous "bridges" are formed.

 

 
 

CONCLUSIONS

IgM was confirmed as the heat-1abi1e component invo1ved
in fat g1obu1e c1ustering. Its participation was a1$o
demonstrated using raw mi1k. IgM was shown to function as
a cryoagqutinin rather than as a cryogTobu1in as previ-
ousTy indicated. The antigen is proposed to be carbohydrate
in nature. SMM was identified as the homogenization-1abi1e
component. A1though IgM can agg1utinate MFG to a 1imited
extent, norma1 creaming requires both components. Approxi-
mate1y 7% of the IgM in norma1 mi1k participates in a sing1e
creaming. The 1ower portion of the creamed mi1k (skim)
fai1$ to support creaming upon addition of washed MFG due
to the absence of functiona1 SMM. The presence of SMM
in the 1ower portion indicates not a11 SMM is capab1e of
participating in creaming. A theory of fat g1obu1e c1us-
tering consistent with the observed experimentaT resu1ts
depicts IgM interacting in an antigen-antibody mode simu1-
taneous1y with SMM and MFG through specific carbohydrate

moieties.

140

 

 

RECOMMENDATIONS

Questions raised by this project and warranting further
investigation are: (a) What is the structure of the MFG/SMM
antigen(s) invoTved in fat g1obu1e c1ustering?, (b) What
specific SMM component (origin) is invoTved?, (c) What is
the mechanism of homogenization-inactivation of SMM?,

(d) ExactTy how does SMM augment creaming - is it mere1y
through its effect on the distance whkm MFG must approach
to a110w c1ustering?, (e) Is a11 of the IgM in mi1k capab1e
of participating in creaming?, (f) What does the tempera-
ture-dependence of the reaction depend upon - the nature of
the antigen, antibody, or reaction conditions?, (9) How
does sa1t addition and pH change effect the two separate
stages of MFG c1ustering?, and (h) What is the structure

of the SMM-IgM-MFG comp1ex?

141

 

 

 

 

 

APPENDICES

 

142

Tab1e A1. ChemicaTS used in this study and their sources.

 

 

Chemica1 Company

Potassium ch1oride A11ied Chemica1

Ammonium persquate J.T. Baker Chemica1 Co.
Potassium carbonate

Acry1amide Bio-Rad Laboratories, Inc.

Bisacry1amide
Sodium dodecy1 su1fate
N-acety1ga1actosamine Ca1biochem-Behring Corp.
N-acety1g1ucosamine
Azoco11
Dithiothreito]
Fucose
Ga1actose
G1ucose
Lactose
Mannose
Xy1ose
High vacuum grease Dow Corning Corp.
Photo-F10 200 Eastman Kodak
N,N,N',N'-Tetramethy1-
ethy1enediamine
Ammonium su1fate Fisher Scientific Co.
BromophenoT bTue
Ca1cium ch1oride
Magnesium citrate
Potassium citrate
Potassium phosphate, monobasic
Potassium su1fate
Sodium citrate

FoIin-Cioca1teu pheno1 Har1eco
reagent

Acetic acid Ma11inckrodt

Barbita1

Ch1oroform

Citric acid

Copper su1fate

Diethy1 ether
Hydroch10ric acid
2-Mercaptoethan01
Methano1

Sodium bromide

Sodium carbonate

Sodium ch1oride

Sodium hydroxide

Sodium phosphate, monobasic
Sodium phosphate, dibasic
Sodium tartrate

Sucrose

 

 

 

143

Tab1e A1. (continued)

 

Chemica1

Company

 

Su1furic acid

TrichToroacetic acid

Ca1cium carbonate

Magnesium carbonate

Methy1 green

ThymoT

Po1yviny1pyrro1idine

Iodoacetamide

B1ue dextran

Low mo1ecu1ar weight SDS
ca1ibration kit

N-acety1neuraminic acid

Bovine serum a1bumin

Coomassie Bri11iant BTue R250

Dowex AGT-XB, 20-50 mesh

Dowex 50W-X8, 20—50 mesh

Ga1actosamine-HC1

G1ucosamine-HC1

Myog1obin

Ova1bumin

Sodium azide

Soybean trypsin inhibitor

Tris (trishydroxymethy1ami-
nomethane)

G1ycine

—\_

MCB Manufacturing Chemists

Oxford Laboratories
Pfa1tz and Bauer, Inc.
Pharmacia Fine Chemica1s

Sigma Chemica1 Co.

U.S. Biochemica1 Corp.

144

Tab1e A2. Equipment routineTy used in this study.

 

Equipment Company

 

Ana1ytica1 ba1ance, type 2463 Satorius Ba1ances
Top-Toading ba1ance, type 3716 Satorius Ba1ances
Top-Toading ba1ance, type K7T Mett1er Instrument Corp.
Camera, MP-3 Land Camera Po1aroid Corp. ,
Preparative centrifuge, mode1 U Internationa1 Equipment Co.
type 277 head
Preparative refrigerated Sorva11 Instruments
centrifuge, mode1 RCZ-B,
type SS-34 and GSA rotors
Preparative refrigerated Beckman Instruments, Inc.
u1tracentrifuge, mode1 L-2-65,
type 21, 30, and 65 rotors
Conductivity meter Industria1 Instruments,Inc.
Visking dia1ysis tubing Union Carbide Corp.
Disk mi1k separator, type LWA Westfa1ia Separator
205

Hand-operated homogenizer C.W. Logeman, Co.
Mechanica1 homogenizer, C.W. Logeman, Co.
mode1 C-8

Ly0phi1izer Laboratory-constructed,
Dr. J.R. Brunner

Convection oven, mode1 OV51O B1ue M E1ectric Co.

Vacuum oven, no. 95050 Centra1 Scientific Co.

pH meter, mode1 245 Instrumentation Labora-
tories Inc.

PH meter, mode1 12 Corning Scientific Instru-
ments

Rotary shaker Eberbach

Water bath, no. C6566 Precision Scientific Co.

Water bath, mode1 RW-150 New Brunswick Scientific Co.

 

 

 

 

 

145

Tab1e A3. Activities of g1ycosidases from Turbo cornutus
as suppTied by Mi1es Laboratories, Inc.

 

G1ycosidase Activity (units)* Substrate**

 

d-Mannosidase 94 PNPG
B-Mannosidase 106 PNPG
d-GTucosidase 3.0 PNPG
B-GTucosidase 7.6 PNPG
q-Ga1actosidase 4O PNPG
B-GaTactosidase 110 PNPG
d—L-Fucosidase 60 PG
o-D-Fucosidase - PNPG
B-Xy1osidase 9.8 PNPG
a-N-Acety1g1ucosaminidase 2.2 PG
B-N-Acety1g1ucosaminidase 460 PNPT
a-N-Acety1ga1actosaminidase 30 PG
B-N-Acety1ga1actosaminidase 44 PG
B-L-Arabinosidase - PG
B-D-Arabinosidase - PG
Neuraminidase - SL

 

*
One unit wi11 re1ease 1.0 umo1e of p-nitropheny1 or pheny1
per minute from gTycoside at pH 4.0 at 37 C.

*9:
PNPG: p-nitropheny1 g1ycoside, PG: pheny1 g1ycoside, SL:
sia1y11actose.

 

 

BIBLIOGRAPHY

 

 

 

BIBLIOGRAPHY

Abo-ETnaga, I.G., G.M. E1-Sadek, and A.M. E1-Sokary. 1966.
C1ustering of fat g1obu1es in cows' and buffa1oes'
mi1k, creaming mechanism and the physica1 arrangement

of g1obu1es in gravity cream. Mi1chwissenschaft 21:
210.

A1der, F.L. 1965. Studies on mouse antibodies. 1. The
response to sheep red ce11s. J. Immuno1. 95:26.

Atherton, H.V., and J.A. New1ander. 1977. Chemistry and
testing of dairy products. 4th ed. AVI Pub1ishing
Co., Inc., Westport, CT.

Babcock, S.M. 1889a. The constitution of mi1k, and some
of the conditions which affect the separation of cream.
Wis. Agr. Exp. Sta. Res. Bu11. No. 18.

Babcock, S.M. 1889b. Fibrin in mi1k. Wis. Agr. Exp.
Sta. Ann. Rpt. 6:63.

Babcock, S.M., and H.L. Russe11. 1896a. Conditions
affecting the consistency of mi1k. Wis. Agr. Exp.
Sta. Ann. Rpt. 13:73.

Babcock, S.M., and H.L. Russe11. 1896b. On the restora-
tion of the consistency of pasteurized mi1k and cream.
Wis. Agr. Exp. Sta. Ann. Rpt. 13:81.

Bea1e, 0., and N. Buttress. 1972. Structura1 studies on
bovine immunog1obu1in M. Biochim. Biophys. Acta 257:
372.

Bertoni, G., V. Cappa, and A. Speroni. 1979. Pre1iminary
study of factors associated with production of mi1k
with a 10w creaming capacity. Sci. Tecn. Latt.-cas.
30:343. (Origina1 not seen. Cited from Dairy Sci.
Abstr. 42:357 (1980)).

B1anc, B. 1979. Biochemica1 aspects of human mi1k —

comparison with bovine mi1k. Page 1 in W1d. Rev. Nutr.
Diet. G.H. Bourne, ed. S. Karger, Base1, Germany.

146

 

147

Bottazzi, V., F. De11ag1io, and P.G. Sara. 1972. Creaming
of mi1k and bacteria1 agg1utination of microorganisms.
IV. Mechanism causing fat and bacteria agg1utination.
Sci. Tecn. Latt.-cas. 23:287. (Origina1 not seen. Cited
from Dairy Sci. Abstr. 352313 (1973)).

Bottazzi, V., and L. Premi. 1977. Re1ation between 5'-

nuc1eotidase and fat g1obu1e agg1utination. Sci. Tecn.
Latt.-cas. 28:7.

Bottazzi, V., and C. Zacconi. 1980. Iso1ation and preTimi-
nary characterization of the active princip1e invo1ved
in aggregation of fat gIobu1es and bacteria. Sci.

Tecn. Latt.-cas. 31:379. (Origina1 not seen. Cited
from Dairy Sci. Abstr. 43:847 (1981)).

Brouet, J.C., J.P. C1auve1, F. Danon, M. K1ein, and M.
Se1igmann. 1974. Bio1ogic and c1inica1 significance
of cryog1obu1ins. Am. J. Med. 37:775.

Brouwer, E. 1924. Over het wezen der vetbo11e1jesagg1u-
tinatie. IV. Over serumg1obu1ine in verband met de
me1koprooming. Proefzuive1boerderij Hoorn 1924:18.
(Origina1 not seen. Cited by Dunk1ey, W.L., and H.H.
Sommer. 1944. Wis. Agr. Exp. Sta. Res. Bu11. No.
151 .

Brunner, J.R. 1974. Physica1 equi1ibria in mi1k: the
1ipid phase. Page 474 in Fundamenta1s of dairy chemis—
try. B.H. Webb, A.H. Johnson, and J.A. A1ford, ed.

AVI Pub1ishing Co., Inc., Westport, CT.

Brunner, J.R., and M.P. Thompson. 1959. Some characteris—
tics of the g1yco-macropeptide of casein - a product
of the primary rennin action. J. Dairy Sci. 42:1881.

Brunngraber, E.G., G. Tettamanti, and B. Berra. 1976.
Extraction and ana1ysis of materia1s containing 1ipid-
bound sia1ic acid. Page 159 in G1yco1ipid methodo1ogy.
L.A. Witting, ed. American Oi1 Chemists Society,
Champaign, IL.

But1er, J.E. 1969. Bovine immunog1obu1ins. A review.
J. Dairy Sci. 52:1895.

Ceriani, R.L., K. Thompson, J.A. Petterson, and S. Abraham.
1977. Surface differentiation antigens of human
mammary epitheTia1 ce11s carried on the human mi1k fat
g1obu1e. Proc. Nat1. Acad. Sci. USA 74:582.

148

Curtain, C.C. 1969. Anti-I agg1utinins in non-human sera.
Vox. Sang. 16:161.

Dah1berg, A.C., and J.C. Marquardt. 1929. The creaming of
raw and pasteurized mi1k. N.Y. State Agr. Exp. Sta.
Tech. Bu11. No. 157. (Origina1 not seen. Cited by
Dunk1ey, W.L., and H.H. Sommer. 1944. Wis. Agr. Exp.
Sta. Res. Bu11. No. 151).

Dah1e, 0.0., and E.L. Jack. 1937. The e1ectrokinetic
potentia1 of mi1k fat. II. Re1ation to dairy processes.
J. Dairy Sci. 20:605.

Davis, B.J. 1964. Disc e1ectrophoresis. II. Method and
app1ication to human serum proteins. Ann. N.Y. Acad.
Sci. 121:404.

Dunk1ey, W.L., and H.H. Sommer. 1944. The creaming of
mi1k. Wis. Agr. Exp. Sta. Res. Bu11. No. 151.

Fahmi, A.H. 1951. The rate of creaming of raw cows' and
buffa1oes' mi1k. J. Dairy Res. 18:106.

Fahmi, A.H., I. Sirry, and A. Safwat. 1956a. Effect of
temperature and time on the creaming of cow, buffa1oe,
sheep, and goat mi1k. Indian J. Dairy Sci. 9:80.

Fahmi, A.H., I. Sirry, and A. Safwat. 1956b. The size
of fat gIobu1es and the creaming power of cow, buffa1oe,
sheep, and goat mi1k. Indian J. Dairy Sci. 9:124.

Farrar, G.H., and R. Harrison. 1978. Iso1ation and struc—
tura1 characterization of a1ka1i-1abi1e o1igosaccharides
from bovine mi1k-fat-g1obu1e membrane. Biochem. J. 171:
549.

Franzen, R.W. 1971. Iso1ation and characterization of
bovine 1actea1 immunog1obu1ins. Ph.D. Thesis. Michi-
gan State University.

Gammack, 0.8., and B.B. Gupta. 1970. C1ustering of mi1k
fat g1obu1es. 18th Intern. Dairy Congr., Proc.,
Me1bourne, 1E:20.

G1ockner, W.M., R.A. Newman, W. Dahr, and G. Uh1enbruck.
1976. A1ka1i-1abi1e oTigosaccharides from g1ycopro-
teins of different erythrocyte and mi1k fat g1obu1e
membranes. Biochim. Biophys. Acta 433:402.

Gonza1es-Jano1ino, V.T. 1968. The creaming properties of
carabaos' mi1k. Part II. A study of the factors inf1u-
necing the creaming abi1ity of carabaos' mi1k.
Mi1chwissenschaft 23:204.

 

 

0
0
0
0

149

Goudemand, M., and Y. De1mas-Marsa1et. 1975. Princip1es
of immunohemato1ogy. Ann Arbor Science Pub1ishers,
Inc., Ann Arbor, MI.

Hammer, B.W. 1916. Studies on the creaming abi1ity of
mi1k. Iowa Agr. Exp. Sta. Res. Bu11. No. 31.

Hansson, E. 1949. The creaming of mi1k. 12th Intern.
Dairy Congr., Proc., Stockho1m, 2:30.

Harrison, R., J.D. Higginbotham, and R. Newman. 1975.
Sia1og1ycopeptides from bovine mi1k fat gTobu1e mem-
brane. Biochim. Biophys. Acta 389:449.

Hekma, E. 1922. Is fibrine een physioTogisch me1kvestand-
dee1? Proefzuive1boerderij Hoorn 1922:1. (Origina1
not seen. Cited by Dunk1ey, W.L., and H.H. Sommer.
1944. Wis. Agr. Exp. Sta. Res. Bu11. No. 151).

Hekma, E., and H.A. Sirks. 1923. Over het wezen der
vetboT1e1jesagg1utinatie. II. Voortgezet quantitatief
onderzoek ontrent den inv1oed b1oedserum en van verhit
b1oedserum op de oprooming. Proefzuive1boerderij Hoorn
1923:88. (Origina1 not seen. Cited by Dunk1ey, W.L.,
and H.H. Sommer. 1944. Wis. Agr. Exp. Sta. Res.

Bu11. No. 151).

Henseva1, M. 1902. Study of rapid and s1ow creaming mi1ks.
Rev. Gen. du Lait 1:366. (Origina1 not seen. Cited
by Hammer, B.W. 1916. Iowa Agr. Exp. Sta. Res. Bu11.
No. 31 .

Hera1d, C.T., and J.R. Brunner. 1957. The fat-g1obu1e
membrane of norma1 cow's mi1k. I. The iso1ation and
characteristics of two membrane-protein fractions. J.
Dairy Sci. 40:948.

Hunziker, O.F. 1921. Pasteurization and the cream Tine.
Hoard's Dairyman 61:91.

Jackson, R.H., and M.J. Pa11ansch. 1961. Inf1uence of
mi1k proteins on interfacia1 tension between butter oi1

and various aqueous phases. J. Agric. Food Chem. 9:
424.

Jenness, R., and J. Koops. 1962. Preparation and proper-
ties of a sa1t so1ution which simu1ates mi1k U1trafi1-
trate. Neth. Mi1k Dairy J. 16:153.

 

 

150

Jenness, R., and S. Parkash. 1971. Lack of a fat g1obu1e
c1ustering agent in goats' mi1k. J. Diary Sci. 54:123.

Jenness, R., and S. Patton. 1959. Princip1es of dairy
chemistry. John Wi1ey and Sons, Inc., New York, N.Y.

Kabat, E.A. 1976. Structura1 concepts in immuno1ogy and
immunochemistry. 2nd ed. Ho1t, Rinehart, and Winston,
New York, N.Y.

Keenan, T.W. 1974. Composition and synthesis of gang1io—
sides in mammary g1and and mi1k of the bovine.
Biochim. Biophys.Acta 337:255.

Kenyon, A.J., and R. Jenness. 1958. Effect of proteins on
agg1utination of fat g1obu1es. J. Dairy Sci. 41:
716.

Kenyon, A.J., R. Jenness, and R.K. Anderson. 1966. Ro1e
of mi1k immunog1obu1ins in fat g1obu1e c1ustering and
the ring test phenomenon. J. Dairy Sci. 49:1144.

Kim, Y.B., S.G. Brad1ey, and D.W. Watson. 1966. Ontogeny
of the immune response. II. Characterization of
195 yG— and 75 y G-immunog1obu1ins in the true primary
and secondary responses in pig1ets. J. Immuno1. 97:
189.

Kitchen, B.J. 1974. A comparison of the properties of
membranes iso1ated from bovine skim mi1k and cream.
Biochim. Biophys. Acta 356:257.

Koops, J., T.A.J. Payens, and M.F. Kerkhof Mogot. 1966.
The effect of homogenization on the spontaneous
creaming of mi1k. Neth. Mi1k Dairy J. 20:296.

Kumar, S., and E.M. MikoTajcik. 1972. E1ectrophoretic,
immunoe1ectrophoretic, and u1tracentrifuga1 character-
ization of proteins in whey fractions prepared by sa1t
fractionation. J. Dairy Sci. 55:1237.

Mancini, G., A.O. Carbonara, and J.F. Heremans. 1965.
Immunochemica1 quantitation of antigens by sing1e
radia1 diffusion. Immunochem. 2:235.

Marcas, L. 1903. A contribution to the study of s1ow1y
creaming mi1ks. Bu11. Agr. (Brusse1s) 19:1228.
(Origina1 not seen. Cited by Hammer, B.W. 1916.
Iowa Agr. Exp. Sta. Res. Bu11. No. 31).

 

.

a, 2..

- u I ..I
..00 to _ ‘
v. . r .l Ink V . on!
_ ._ 0... .. ..k. .
..C ‘I v 0 h .v n

.
l

 

151

Marcas, L. 1904. Contribution to the study of s1ow1y
creaming mi1ks. Rev. Gen. Lait 3:361. (Origina1 not
seen. Cited by Hammer, B.W. 1916. Iowa Agr. Exp.
Sta. Res. Bu11. No. 31).

Markwe11, M.A.K., S.M. Haas, L.L. Bieber, and N.E. To1bert.
1978. A modification of the Lowry procedure to
simpTify protein determination in membrane and 1ipo-
protein samp1es. Ana1. Biochem. 87:206.

Mather, I.H., and T.W. Keenan. 1975. Studies on the
structure of mi1k fat g1obu1e membrane. J. Membrane
Bio1. 21:65.

Mckay, G.L., and C. Larsen. 1906. Princip1es and practice
of butter-making. John Wi1ey and Sons, New York, N.Y.
(Origina1 not seen. Cited by Dunk1ey, W.L., and H.H.
Sommer. 1944. Wis. Agr. Exp. Sta. Res. Bu11. No.

151 .

Me1achouris, N. 1969. Discontinuous ge1 e1ectrophoresis
of whey proteins, casein, and c1otting enzymes. J.
Dairy Sci. 52:456.

Mertens, E. 1932. Beitage zum aufrahmungsprob1em. I.
Praktische versuche zum prob1em der aufrahmung.
Mi1chw. Forsch. 14:1. (Origina1 not seen. Cited by
Dunk1ey, W.L., and H.H. Sommer. 1944. Wis. Agr.
Exp. Sta. Res. Bu11. No. 151).

Metzger, H. 1970. Structure and function of IgM macro-
g10bu1ins. Page 57 in Adv. in ImmunoT. F.J. Dixon,
Jr., and H.G. Kunke1, ed. Academic Press, New York,
N.Y.

Mi1grom, F. 1969. Autoimmunity. Vox. Sang. 16:286.

Mukkur, T.K.S., and A. Froese. 1971. IsoTation and
characterization of IgM from bovine co1ostra1 whey.
Immunochem. 8:257.

Mu1der, H., and P. Wa1stra. 1974. The mi1k fat g1obu1e;
emu1$ion science as app1ied to mi1k products and com-
parab1e foods. Pudoc, Wageningen and CAB, Farnham
Roya1, Eng1and.

Newman, R.A., and R. Harrison. 1973. Characterization of
the surface of bovine mi1k fat g1obu1e membrane using
microe1ectrophoresis. Biochim. Biophys. Acta 298:798.

152

Newman, R.A., R. Harrison, and G. Uh1enbruck. 1976.
A1ka1i-1abi1e o1igosaccharides from bovine mi1k fat
g1obu1e membrane g1ycoprotein. Biochim. Biophys.
Acta 433:344.

Newman, R.A., and G. Uh1enbruck. 1977. Investigation into
the occurrence and structure of Tectin receptors on
human and bovine erythrocyte,_mi1k-fat g1obu1e and
1ymphocyte p1asma-membrane g1ycoproteins. Eur. J.
Biochem. 76:149.

Or1a-Jensen, S. 1929. Recherches sur 1a pasteurisation
basse dur 1ait. Une nouve11e reaction pour contro1er
1a temperature da chauffage. Lait 9:622. (Origina1
not seen. Cited by Dunk1ey, W.L., and H.H. Sommer.
1944. Wis. Agr. Exp. Sta. Res. Bu11. No. 151).

Ouchter1ony, O. 1949. Antigen-antibody reactions in ge1s.
Acta Patho1. Microbio1. Scand. 25:507.

Pa1mer, L.S., and E.O. Anderson. 1926. Physicochemica1
factors inf1uencing cream rising. I. Viscosity.
J. Dairy Sci. 9:1.

Pa1mer, L.S., J.C. Hening, and E.O. Anderson. 1926.
Physicochemica1 factors inf1uencing cream rising.
II. Re1ation of p1asma co110ids to pasteurization
effects. J. Dairy Sci. 9:171.

Patton, S., E.R. Bogus, B.H. Sternberger, and E.G. Trams.
1980. Antiserum to the mi1k fat g1obu1e membrane-
preparation and capacity to suppress mi1k secretion.
Biochim. Biophys. Acta 597:216.

Patton, 3., and T.W. Keenan. 1975. The mi1k fat g1obu1e
membrane. Biochim. Biophys. Acta 4152273.

Payens, T.A.J. 1964. Das prob1em der ko110idchemischen
stabi1itaet der fettkuegeTchen in der mi1ch.
Kie1. Mi1chwirtsch. Forschungsber. 16:457.

Payens, T.A.J. 1968. Neure forschungen auf dem gebeit
der Mi1cheiweisse. Mi1chwissenschaft 23:325.

Payens, T.A.J. and P. Both. 1970. Cryog1obu1ins from mi1k
and a mechanism for the co1d agg1utination of mi1k-fat
gTobu1es. Immunochem. 7:869.

Payens, T.A.J., J. Koops, and M.F. Kerkhof Mogot. 1965.
Adsorption of eug1obu1in on agg1utinating fat g1obu1es.
Biochim. Biophys. Acta 94:576.

 

 

153

P1antz, P.E., S. Patton, and T.W. Keenan. 1973. Further
evidence of p1asma membrane materia1 in skim mi1k. J.
Dairy Sci. 56:978.

Po11ack, W. 1965. Some physicochemica1 aspects of hemagg1u-
tination. Anna1s N.Y. Acad. Sci. 127:892.

Po11ack. W., H.J. Hager, R. Recke1, D.A. Toren, and H.O.
Singher. 1965. A study of the forces invo1ved in the
second stage of hemagg1utination. Transfusion 5:158.

Race, R.R., and R. Sanger. 1968. B1ood groups in man.
5th ed. F.A. Davis Co., Phi1ade1phia, PA.

Rahn, O. 1921. Untersuchungen uber die Rahmbi1dung.
Forsch. geb. Mi1chwirtsch. u. Mo1kereiwesen i:133.
(Origina1 not seen. Cited by Troy, H.C., and P.F.
Sharp. 1928. J. Dairy Sci. 11:189).

Reisfe1d, R.A., U.J. Lewis, and D.E. Wi11iams. 1962.
Disk e1ectrophoresis of basic proteins and peptides
on po1yacry1amide ge1s. Nature 195:281.

Reisfe1d, R.A., and P.A. Sma11, Jr. 1966. E1ectrophoretic
heterogeneity of po1ypeptide chains of specific anti-
bodies. Science 152 1253.

Rhee, K.C. 1969. Iso1ation and characterization of 1actea1
immunog1obu1ins. Ph.D. Thesis. Michigan State Univer-
sity.

Rodbard, D., and A. Chrambach. 1971. Estimation of
mo1ecu1ar radius, free mobi1ity, and va1ence using
po1yacry1amide ge1 e1ectrophoresis. Ana1. Biochem. 40:

Row1and, S.J. 1937. The creaming power of heated mi1k.
The re1ationship between the denaturation of a1bumin
and g1obu1in and the reduction in creaming power.

J. Dairy Res. 8:195.

Samue1sson, E., G. Bengston, S. Ni1sson, and N. Mattsson.
1954. Cream rising. Svenska Mejeritidn. 46:163.
(Origina1 not seen. Cited from Dairy Sci. Abstr. 16:
1015 (1954)).

Scott, D.W., and R.K. Gershon. 1970. Determination of
tota1 and mercaptoethano1-resistant antibody in the
same serum samp1e. C1in. Exp. Immuno1. 6:313.

 

 

 

154

Segrest, J.P., and R.L. Jackson. 1972. Mo1ecu1ar weight
determination of g1ycoproteins by po1yacry1amide ge1
e1ectrophoresis in sodium dodecy1 su1fate. Page 54
jfl_Methods hienzymo1ogy. V01. 28. V. Ginsburg, ed.
Academic Press, New York, N.Y.

Sharp, P.F., and V.N. Krukovsky. 1939. Differences in
adsorption of so1id and 1iquid fat g1obu1es as inf1uen-

cing the surface tension and creaming of mi1k. J.
Dairy Sci. 22:743.

Shimizu, M., K. Yamauchi, and C. Kanno. 1979. Proteo1ytic
digestion of mi1k fat g1obu1e membrane proteins.
Mi1chwissenschaft 34:666.

Shimizu, M., K. Yamauchi, and C. Kanno. 1980. Effect of
proteo1ic of mi1k fat g1obu1e membrane proteins on
stabi1ity of the g1obu1es. Mi1chwissenschaft 35:9.

Smith, E.L. 1946a. The immune proteins of bovine co1ostrum
and p1asma. J. Bio1. Chem. 164:345.

Smith, E.L. 1946b. Iso1ation and properties of immune
1actog1obu1ins from bovine whey. J. Bio1. Chem. 165:
665.

Stadhouders, J., and G. Hup. 1970. Comp1exity and speci-
ficity of eug1obu1in in re1ation to inhibition of
bacteria and to cream rising. Neth. Mi1k Dairy J. 24:
79.

Stone, J.S., and H. Metzger. 1968. Binding properties of
a Wa1denstrom macrogTobu1in antibody. J. Bio1. Chem.
243:5977.

Tomich, J.M., I.A. Mather, and T.W. Keenan. 1976. Proteins
.mask gang1iosides in mi1k fat g1obu1e and erythrocyte
membranes. Biochim. Biophys. Acta 433:357.

Troy, H.C., and P.F. Sharp. 1928. Physica1 factors inf1u-
encing the formation and fat content of gravity cream.
J. Dairy Sci. 11:189.

VanDam, W., E. Hekma, and H.A. Sirks. 1922. Over het
wezen der vetbo11e1jesagg1utinatie. I. Voor1oopig
kwantitatief onderzoek omtrent den inv10ed van b1oed-
serum van verhit b1oedserum en van onderme1k op de
oprooming. Proefziuve1boerdrij Hoorn 1922:81.
(Origina1 not seen. Cited by Dunk1ey, W.L., and H.H.
Sommer. 1944. Wis. Agr. Exp. Sta. Res. Bu11. No.
151 .

 

 

155

VanDam, W., and H.A. Sirks. 1922. Onderzoekingen over de
oprooming va1gens het Froesch systeem. Vers. 1andb.
Onderzoek Rijs1andbauw. 26:106. (Origina1 not seen.
Cited by Troy, H.C., and P.F. Sharp. 1928. J. Dairy
Sci. 11:189).

Van Der Burg, B. 1921. De beweging van een vetbo11e1je
in de me1k. Het A1gemeen Zui1ve1b1ad. Jan. 7:3.
(Origina1 not seen. Cited by Troy, H.C., and P.F.
Sharp. 1928. J. Dairy Sci. 11:189).

Wa1stra, P. 1980. The effect of homogenization on mi1k
p1asma. Neth. Mi1k Dairy J. 34:181.

Weber, K., and M. Osborn. 1969. The re1iabi1ity of mo1e-
cu1ar weight determinations by dodecy1 su1fate-P01Y-

acry1amide ge1 e1ectrophoresis. J. BioI. Chem. 244:
4406.

Weeke, B. 1973. Genera1 remarks on princip1es, equipment,
reagents, and procedures. Page 15 in A manua1 of
quantitative immunoe1ectrophoresis. N.H. Axe1sen, J.
Kro11, and B. Weeks, ed. Universitatsfor1aget, 0510,
Norway.

Whitney, R. McL., J.R. Brunner, K.E. Ebner, H.M. Farre11,
R.V. Josephson, C.V. Morr, and H.E. Swaisgood. 1976.
Nomenc1ature of the proteins of cow's mi1k: fourth
revision. J. Dairy Sci. 59:795.

Whitt1estone, W.G. 1953. Variations in the fat content of

mi1k throughout the mi1king process. J. Dairy Res.
20:146.

Wiener, A.S., L.J. Unger, L. Cohen, and J. Fe1dman. 1956.
Type—specific co1d auto—antibodies as a cause of
acquired hemoTytic anemia and hemoTytic transfusion
reactions: Bio1ogic test with bovine red ce11s. Ann.
Intern. Med. 44:221.

Wiman, K., B. Curman, L. Tragardh, and P.A. Peterson. 1979.
Demonstration of HLA-DR-1ike antigens on mi1k fat
g1obu1e membranes. Eur. J. Immuno1. 9:190.

W011, F.W., S.M. Babcock, and H.L. Russe11. 1903. Condi-
tions affecting the consistency of mi1k; Means of
restoring consistency of pastuerized cream. Wis. Agr.
Exp. Sta. Ann.Rpt. 20:183.

 

156

Yamauchi, K., M. Shimizu, and C. Kanno. 1978. Effect of
preparation on properties of a soTub1e g1ycoprotein

fraction of mi1k fat g1obu1e membrane. J. Dairy Sci.
61:688.

Zinneman, H.V. 1978. Cryog1obu1ins and pyrog1obu1ins.
Page 323 in Immunog1obu1ins. G.W. Litman and R.A.
Good, ed. P1enum Medica1 Book Co., New York. N.Y.

Zitt1e, C.A., and J.H. Custer. 1963. Purification and
some of the pr0perties of as-casein and K-casein.
J. Dairy Sci. 46:1183.

Zmijewski, C.M., and J.L. F1etcher. 1972. Immunohemato—
1ogy. 2nd ed. App1eton-Century-Crofts, New York,
N.Y.

111111111Ill!11111111111|111111111111111