A STUDY OF THE DETECTION OF COLIFORM ORGANISMS SEEDED IN STERILE SKIM MILK Thesis for the Dogma of M. S. MiCHlGAN STATE COLLEGE Manley Mandel W47 IBESVE This is to certify that the thesis entitled ”A Study of the Detection of Coliform Organisms Seeded in’ Sterile Skim Milk" presented by Manley Mandel has been accepted towards fulfillment of the requirements for M._degree inmmlogy fl/W M’ajor prof! sor Date—WW- _ 11-795 i h .I.’ ,.. on 1' J. . .v . 1‘ . .v.’tut . . p '1'! p .1, 0| _ A. can a»? a. .3... n. C . .i \ «anm ., J \- r v . A I IP . T . .0 . u ..-l - ph. r a. s .t : draw? . a‘b ‘ . V f \ . . “th; Au ,‘ . . T.-. . I. .. v. . "W‘IU‘D.3 0.x) ...\ u. ‘ . r. I V... .v . v x v\.\ .w..% .nt .n. ‘ . ‘ . Q I I ‘ I I I It]. ‘I A. b u t I v i - ¢ 1. I u .0 t ‘- . t .— n _ u _ to . e. n _ _ a O I‘ I‘| In! I . .- . . s, , Au . .‘ u o n . v .c . 1 . . . . . y . \‘rl W”. ‘1" o? . It . .\ C st «I ‘v a .—I A. V) . if. u I ~ I r‘ n . . o til i _ »@~ .‘ . o .§ . I e . - v .. l v \ ' 11". . te ’ ‘ s U . . I. . . . . . . .u .. Av . .‘1‘ . .0 I n _ n l O . t . i. , . . 1 .I y .g .. . t . , . cVNDnmi». . .. . . . . ... . . . 1 . .. . . . ‘ ‘\ ... cl . v, . . . a . . o .I f t . . ._....e.-.\ . . “all .. .3“-.. A :1 s ._ 1' .n. 5.x. . u: . _. \. r! 4 ..«.f ‘4 .. . n..V.vJ\I ‘ I - Rm . ii. .3... t . .r h 3 . .3 65%th .23 “Mafia. .- - n bTUDY OF Tun UanCTIUN OF COLIFOhM Unbnmlbmo omnbnu IA bimnlbfi sxln MILK By LLANLnY‘uhQMJEL 11 131113015 Submitted to the bohcol of Graduate studies of Michigan state College of agricultgre and “9911 d science w in partial fulfillment of the reouirements for the degree of unblmn OF oClLde Department of bacteriology 1947 5/ 57/ 53/ The author wishes to eXpress his sincere apprec- iation to Dr. A. L. ficrtree for his interest and guid- ance in these studies, and to Dr. H. J. Stafseth for his assistance in the preparation of this :anuscript. ThBLfl OF CONTBNTS IDtI‘OdUCtiOH 00000.000000000000000...00000000000 Experimental procedure ......................... hQSLlltS OOOOOOOOOOOOOOOOOOOIOOOOOOOOOCOOOOOOOOCO l. Cont rison of plate and microscopic counts or orgunisns seeded in sterile sxim milk . 2. Compsrison of counts when the organisms to seeded were gromn unier various conditions 5. Microscopic epoesrance of the stained Organisnls 0.0.0...OOOOOOOOOOOOOOOOOOOOO... 4. Studies of Stains and staining procedures DichfSSion .0...0.0.00.0...OOOOOOOOOOOOCOOOOCOO bdfiliflary COOOOOOOOOOOOOOOO0.00000000000000000000 fiefer.€fnces .OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO A£)l:’endix . C O C O O C O C C O O O O O O O C O O O C C C O O O Q C O C O O . 0 O O 0 Chain F‘jli'iHUlae . o o o o o o o o o o o o o o o o o o o o o o o o o o o o o pl'OCeil‘e 7‘9 0 o o o o o o o o o o o o o necomnended staininr [iiistlkl‘S 00000000000000000000000000000000.0000. 0e 0 .16 INTHOUUCTION The Breed direct microscopic examination of milk (5) has been widely usei For the enumeration 3f bacteria in milk. Numerous efforts heve been male to compare the re— sults obtained by the direct microscooic methol and those obtained by the agar nlate method (b,o,14,26). Most workers have found that the tto methods are not directly comparable (o,7,l*,25,so). an excellent discussion of the relationship of these two procedures has been presen— ted by Hammer (lo). Duplicate Samples of milk counted by these two methods have shown, in most cases, that the mi- croscopic count is far higher than the plate count. hobertson (20) reported that the mode was 4:1 when the counts were compared, direct microscopic examination to agar plates incubated at 6700 for 45 hours. exceptions to this relationship are to be found in data presented in the literature there the two methois h:ve been com- pared (1,2,o,1e,1o,17,1e,27). In these comoarisons, nu- merous instances were encountered where the standard slate count was significantly higher than the count ob- tained by direct microscooic examination. Baker (1), in his investigation of the direct micros- copic method for counting bacteria in pasteurized milk, notes the variation in the ability of different species of bacteria to stain with methylene blue. Various stains and staining nroceiures have been recommended to enhance the she d or accuracv of the direct microscopic count. The Newman-Lambert stain (25), drb's rabid stain (13), and Shutt's ranii staining method (54) were designed to de- crease the time necessary for oreoaration of the milk films for examination. Mallmann's acid stain (19), de- visei far the examination of liquid egg meats has been used for the examination of milk (6). In the examination of dunlicate samples this procedure demonstrated counts higher than those obtained by the Breed technique (17). The search f>r better staining procedures since then has centemwvi on the :nu3 of Mallnuuro's metfixxi of backgnmnuri de- colorization. In lflol, hhitehead (2”) rerortei a phenomenon Which is the :aison d; Stre of this investigation. Whiteheai grew various pure cultures on meat—extract neotone Hgar for 24 hours. The organisms were washed from the agar with sterile saline solution ani se-ded into milk as a cheese starter. when the milk sambles so seeded were immediately examined for bacterial content, it was found that the staphylococci, strentococci ani gran—positive bacilli could be easily counted by the direct micros oric methoi. The gram—negative coniform organisms seeded in the milk couli not be detected by iirect microsconic examination. These oréanisms could be shown to be nresent in large numbers by means of a sub- culture dilution technique emoloving MacConkeV's broth. Whitehead investigated this phenomenon further and found it to be consistent. In addition, the temoerature of in- cubation was shown to affect the length of incubation time required for the organisms to become visible upon direct microscopic examination. In this laboratory, the results obtained by hhitehead have been confirmed (o) and attempts at explanation assayed. Two main possibilities exist. The first, advanced by Baker and Peters (2), concerns the possibility of inadequacy of the staining technique employed by whitehead and that used in this laboratory. The second, the possibility that a hitherto undiscovered life cycle of coliform bacteria may exist. The first anOthLSiS is based on the known variability of stain retention by some bacteria, the difficulty with which blue-stained organisms are seen against a blue back- ground, and the possibility that the surface charge carried by the bacteria affects the absorption of the methylene blue dye. The second hypothesis is based on the discovery, by eudbe (2;), of a life cycle of a thermophilic organism found in milk. The organism studied by nudge had an agparent 5 to 8 minute generation time. Further studies revealed that the initial count was loo beCause the spores failed to re— tain the stain and that the rapid increases in counts in the early stages of incubation were due to the staining of the germinated spores in adiition to the normal vegetative reproduction of the organism. since the incubation of milx samples seeded with coliforms showed a similar tremendous increase in numbers when followed microscopically (0,27) during the first hour of incubation, it was reasoned that a similar cycle may be involved in this phenomenon as in that investigated by nudge. The purpose of this investigation, then, is to attempt to confirm the observations of'whitehead, using more rigid conditions, and to inves*igate the two hypotheses propounded above as a means of explanation of the failure of the direct microscopic method to detect coliform organisms seeded in milk. The significance of this situation is readily apparent. A group of organisms of public health significance appear not to be detected by the direct microscopic examination of raw milk. Certainly the presence of numbers of coliform organisms in raw milk is equally as significant as the num- bers of other saprobhytic organisms found in raw milk. More- over, the coliform organisms in raw milk may not have the same significance as tiose found in pasteurized milk, but may be an index of fecal contamination of the raw milk supply as well as of utensil contamination. .L anunlmmmfaL rnanutnm Laboratory strains of escherichia coli, aerobacter aerogenes, an lnt-ruedi,oe coliform and micrococcus caseolyticus were creczed for purity and grown on nutri- ent agar slopes. Incubation was at oVOC for stated peridds. The culture used was washed off the agar slope with 2 ml. sterile skim milk, shaken and seeded into a tube of sterile SKih milk. The dilution so made was im— medi tely shaken and samples witndéawn for counting. stan- dard aéa plate counts were made by the method described in standard Methods for the examination of hairy Broducts (25). Duplicate smears were made for the direct microscopic examination by the method of bryan, Mailmann, and Turney (a). as many slides were made as were required for the different staining procedures. The seeded milk sample was incubated at o7OC and remived at fixed intervals for the preparation of plates or milk films as described above. The sterile skim milk was prepared from low count rat milk obtained at the Michigan state ColleLe Dairy. The skimmed milk was tvbed, steeued one hour on three suc- cessive days and incubated at room temperature between treatments. after sterilization, the milk was checked for {K “tine Q C sterility by incub and by plating. The milk prepared in this mxnner was free from viable organisms and from detectable dead organisms retaining the methylene blue stain. The methylene blue stain routinely used in this laboratory is a modification of the classical creed stain. The formulae of the methylene blue stains used LT in his invrstigation are found on gage i of th appendix. The ph of the Bread and Ceepryn-methylene blue stains was 0.0 to 7.0 with no adjustment. In counting the bacteria in the mila films, as fields and in many cases so fields were counted in duplicate pre- parations. where counts were exceedingly high, only 5 fielis were examined. The counts reported regresent the numbers of individual cells or clumps of cells retaining the stain. Organisms, grown in nutrient broth and in sterile skim milk for different lengths of time, were also treated in like manner to determine if the phenomenon under inves- tigation was also evident with cells grown on media other than agar. The specific conlitions are detailed with each experiment under haJULTd. nnoULTo 1. Comparison of plate and iirect microscopic counts of aninms seeded in sterile skim milx. The data in Table I are the results of counting the bacteria in milk after seeding as described in the experimental procedure. The counts were made, in dupli- cate, on samples withdrawn initially and at intervals of at, so, 90, and 12u minutes incubation at oVOC. No 9b minute sample of the milk seeded with m. caseolyticus was withdrawn. The data, representing the arithmetic mean of the duplicate covnts, confirm whitehead's observations. The initial plate counts are approximately 15 to be times as great as the direct microscopic counts where the coliform organisms have been seeded into sterile skim milk. After 90 to 12b minutes incubation, the counts by the two methods are an roximately in 1:1 ratio. The sample seeded with a. caseolyticns yielded counts in approximately a 1:1 ratio at all times. The ranhe of experinental error between the two technicues is illustrated by the latter set of data, and is approximately Zap. In addition, the data for the coliform organisms durinb the first 90 minutes in the milk denonstrate that the bacteria are not yet in the logarithmic growth phase, when followed by tie agar plate method. The counts ob- tained by the direct microscopic method, however, Mould iniicste that t”e organisms are multiplying logarithmi- cally from the time of inoculation. The data presented in Table II further illustrate the above results. These results were obtained in the same manner as the preceeiing data. When the samples mere withdrawn inmeiiately after seedint with the coliform organisms, the plate Counts were from 15 to low times as high as the direct microscopic counts of each sample. after ob minutes inCLbation of each sample at 0700, the counts are in Closer awreement. The data for sterile Skim milk seeded with M. caseolyticus are included for compari- 83H. 2. Comparison of direct microscopic counts when the O—t ,_) organisms to be SrtdEd were grown on agar sloyes, in nutrie broth, and in skim milk for different periods. The data in Table III were gathered in the same manner as the grevious data, but no plate counts were obtained. The results show that when the organisms were grown in nutrient broth for 0 hours at 0700 and then seeded in sterile skim milk, the direct counts obtained immediately were not significantly lower than those obtained after the samples were incubated for oo minutes at 0700. The Intermediate coliform showed a 2.o—fold increase in that period, which would indicate that no lag was taking lace ’1; under the conditions of transfer to the new medium. The direct counts on milk Samples inoculated from 34 and 46 hour agar slope Cultures incubated at 0700 (J preSAnt a different picture. after oo minutes incu ation ‘ e x _ . o 1 _ _ of tre seeded milk samples at o? C, tnere is evident a o to bo-fold increase in the direct counts. Such increases denote a generation time of a to 14 minutes. When the fact is considered that tie cultures should still be in -9- the l L phase, it must be realized that some factor other hen reproiu tion is involVed. The data presented in Table IV confirm these ob- 0" servations and the plate coun,‘ on each Sample are also presented. The Q hour broth culture has been replaced by a 5 hour broth culture, and 24 hour cultures in nut— rient broth and on agar slopes compared as sources of inocule. The microscopic counts for the samples seeded with the Intermediate coliforu rom the b hour broth culture again show a large increase in the 1 hour incu- bation at oVOC, anl this increase is paralleled in the plate counts. This observation confirms the previous postulation that this organism has no lag phase when transferred under the conditions of this experiment. additional data, not presented in the table, when the counts were taken over shorter intervals bear out this assumption. The bacterial counts of the milk samples seeded from 24 hour broth cultures and 89 hour agar slope cul- tures repeat the patterns sho.n before. Initially, the microscopic counts are low and the plate counts high. after UV minutes incubation at o7OC, the microscopic counts show considerable inoreases, brinéinb them into closer cojrelation with the plate counts. The Intermed— iate coliform inoculated from a 84 hour broth culture showed one set of counts discrepant from the others. A second set obtained in like manner, did not show this discrepancy. -10- A 24 hour culture of g. coli grotn on an agar slope at 5700 was seeled in sterile skim milK. The direct microSCOpic count was e2 thousand and the plate count was loco thousand. The sample was inCubated at 0700 and 1 ml samples withirawn after 90 minute, 0 hour and o hour incubation periods and seeded into tubes of sterile ssim nilx. The direct count on the sample seeded after 9o minutes was 172 thousand and the plate count 170 thousand. after 3 hours the direct counts on the milx sample thus se:ded were oo thousand and the plate count was 8U thousand. The sasple seedel with the organisms incubated for o hnurs gave a direct count of 5400 thou- sand and a plate count of 44UU thousand. It was evideit that the fresh nilx medium was not influencing lacx of correlation between counting methods, but that tee age of the organisms seeded in the inm milK determined the correlation of tne direct counts to the slate counts. The results at this point in tne investigation show that culiform organisms seeded into sterile sxim milk fPOm 94 hour cultures in brsth or on atar do not agpear on im- mediate micrvscoyical examination of the seeded sterile skim milk. After incubation for periods of co minutes or longer, the organisms may be found by direct microscopic examination in the numbers indicated by plate counts. The age of the cultures ns~d for seedint and not the medium -11- upon which the organisms have been grown appears to decide whether the phenomenon will be demonstrated. In the r mainder of this investigation, all parent cultures were grown on nutrient agar lepes for 24 hours at 0700. These cultures are easy to hanile and maintain and demon- strate the phenomenon best when seeded in the milk. 0. Microscopic appearance of the stained organisms. .L The folio ing statements hold for all the coliform strains used in this study. The organisms, when initially seeded in the skim milk and examined after staining with the breed stain, are extremely small and retain the stain poorly. The length of individual cells of g. poll in milk immediately after seeding from a 2a hour agar lepe culture was measured, using a binocular microscope with a 1.8 mm. objective, a 10X ocular for locating the organisms, and a Leitz Wetzlar measuring vernier. The average length of the cells was 1.201p, with variation between the extremes of 0.5'r2.uo,u. Careful examination, with 10X oculars in place of the ox oculurs used in making the counts, revealed "ghost" cells which were not stained with methylene blue. On incubation, the cells increased considerably in length. examination at so and do minutes after inocula- tion revealed larger cells in which the staining could be observed more readily. The most noticeable charac- teristic of the poorly stained organisms was the bizarre manner in "hich the stain was absorbed. The cells stained as 1g common in the genus Corynebacteriuh in many in- stances, that is a barred appearance was noted. In some instances, a bipolar staining was found. The larger cells, in active renroduction, retained the stain uniformly throughout their length. The irregular staining was ob- served only in the smaller cells present at the behinning of incubation. The cells measurei at 9; minutes had an average length of 2.oB,u, betmecn the limits of 1.4e-o.26,u. The cells measurei at lJU minutes had an average length of o.72,u, between the limits of 2.64-5.28,u, Cells measurei at etc minutes averaged 1.59,u, between the limits of 1.2U- 1.92/U. The latter cells were r";idly dividing, and al— though small, retainel toe stain very well. Little or no clumping we evident in any of the shears examined, and could not account for the low micros- coric counts. No spores were encountered at any tine, nor any body yhich might be construed to be a spore. BeCaLse of the evidence presented to this point, namely that olier cells appeared to be responsible for this phenonenon, that these ”physiologically old" cells retained the l tain poorly or not at all, and that no spores were (0 found in the milK films, it was decided to discard the hypothesis that a life cycle of the coliforu organisms -10- was involVed in the phenomenon bein; investigated. The working hypothesis followed from this point on, was that the organisms failed to absoro the methylene blue stain to any great extent, and that the greatest numbers failed to stain at all when freshly seeded into sterile akih milk from a 24 hour agar slape culture. The remainder of the investigation therefire deals with hethdds desiLned to stain the organisns failing to stain with the Breed stain. 4. studies of various stains and staining procedures. as nQLed in the introduction, Mdllmann's acid stain was the only grocedure shown to stain organisms in milk which the Breed technique failed to disclose (lo). Twelve replicate slides were prepared from sterile skim milk seeded with 24 hour agar slope cultures of E. coli and the Intermediate coliform. Four smears of each milk sample were prepared in this manner. all sliies were dried, fixed in heat, defatted 1 minute in xylene fol- lowed by 1 minute in 95 percent ethanol. One slide was stained with the Breed stain and another with hallmann's acid stain, counted and the results compared. The data, in the first row in Table V labeled "nil", show direct correlation. The slight increase in each count with l Mallmann's stain is insignificant . Significance, when used in comparing counts in this study, does not have the eXact meaning as used in statistical an- alysis. In low direct counts, 8 difference of one bacterium between two counts may appear as a 2-fold increase. Likewise, counts of lB,Ucu and oo,be must be considered as not signi- ficantly different. If the true number present were 24,9UU, -14- hec:ntly,Dyar (lo) has demonstrated the use of a cationic surface-active agent as a moriant. In addition to changing the charge of the cell Wall, penetration of the cell membrane is demonstrated, and a fat dye will non stain the cytOplasm, whereas it failed to stain be- fore treatment with the surface-active afient. To determine whetder increased counts could be ob— tained when tVe stains were used in conjunction with surface-active agents, the_following procedure was performed. The remaining lo slides, referred to above, were used. approximate c.01 molar aqueous solutions of alkyl dimethyl benzyl amhonium chloride, Tergitol 4, Tergitol 7, cetyl pyridinium chloride, and Uuponal L-lel WET? preoared. Three drogs of each abent were added to each smear on duhlicate slides. One set of slides was then flooded with the Breed stain and the analOEOus set with hallmann's acid stain. The exposure to tke breed stain was 1 minute and to hallmann's stain 4 minutes. All sliies were briefly immersed in a beaker of distilled water to renove the excess stain and air dried. The data in Table V present the avera es of 4 smears counted when stained in th's manner. The siénificant k) (cont'd) the difference repreSents olus or minus one bacterium in the microscopic counts. In counts below lUb,UUU a ozl dif- ference or greater has been taken to represent significant diff rences, while in higher counts a 3:1 ratio has been considered a significwnt difference. -15- results are that with the Breed stain, the cationic agent cetyl pyridinium chloride and the anionic Duponal L-ldl gave approximate 10-fold increases in the count over the duplicate SM¢&TS stained tith the Breed stain alone. The Smears treated with t~e cetyl pyridinium chloride were extremely clean and uniform and the background was decolo- rized. aith the sliles treated with uuponal L-lel, there were likewise increased counts, bet the films contained a large number of crystals and some detritus, although there were increases with other aéents, the increases were less sig- nificant. when Mallmann's acid stain folloued the treatment with the surface-active agents, there was considerable loss of the milk film. The film iMmGdidtle floated off the slide ween the stain was added in the instances where loss occurred. The cetyl pyridinium chloride gave the bet results of all the agents tested when follOWed by the breed stain. It was decided to concentrate the investigation on the use of this cationic agent in conjunction with the breed stain. To insure the use of uniform concentrations of stain and cationic agent, toe two were combined and used in a coplin jar, rather than in segarate solutions. The composition of the stain is Liven in the appendix under the name of Ceeprvn-methylene blue stain. -10.. The concentration of the cetyl pyridinium chloride in the Ceenryn-methylene blue stain was varied. Con- centrations of the agent giving final m/bU, M/lUU, and m/aoo solutions in the stain were compared with each other, with slides stained by the Breed technique, and with the staniard plate count. E. coli and fl. caseolgti- ESE were the organisms seeded in the mile for these comparisons. The data sumnarizing these comoarisons are presented in Table VI. The highest counts obtained by direct micrOSCOpic examination were in the smears stained with the Ceepryn-methylene bite stain containing M/lOO cetyl gyridinium chloride. The counts are not comparable to the plate counts, but are significantly higher than the Breed count when E. coli was initially seeded in tve ste-ile skim milk. This preparation did not affect the counts of m. caseolyticus, although the stain containing double the concentration of cetyl pyridinium chloride resulted in a general lowering of the counts. The Ceeprvn-methylene blue stain containing M/lOU cationic agent in o0 percent ethanol was next comeared with a life gregaration where 7U percent eth nol was used in making up to volume. A oo-day old “retaration and a 1-day old xrefiaration of the reLular Geopryn-methb- lene blte stein was used. Direct counts usinb the breed stain and st niard agar plate 0 unts were included in the comparison. The organisms seeded in the ste ile -17- skim milk were the Intermediate coliform and n. aerogenes. The data are oresen5ei in Table VII. Botn Ceepryn- methylene blue sreoarations containing ou percent ethanol and m/luo cetyl pvridinium chloride gave the highest direct counts. Thirty days aginx had not deCreasei the efficiencv of the stain. again, it must be noted that even these inoreased direct counts were not as high as the plate counts on the initial sanples. They are hotever, from 5 to e times greater than the direct counts obtained emgloying the Breed stain or the Ceetryn-dethvlene blue stain containing 7o percent alcohol. at this goint toe Ceeprgn-methylene blue stain con- taininb m/on cetyl pyridinium chloride and ou percent ethanol was accented as being the est combination studied. This preparatiun was used from this point on. The time of luquwflnni'of theluilx films in the Gee ryn- methylene blue stain was the next factor consiiered. le shim milK were seeded with all four Ho bangles of ster OTCfifllSLS stuiied and shears inmeiiatelV made for lirect microscapic examinatian. btaniari agar plates were pre- pared at the sad (D time. The films were fixed in heat, defcttei in xvlene, and rinsed in ab percent alcohol. One slide was stained with dreed's stain for 1 minute. Three slides were stainei in Ceepryn-methylene blue for -18— oo seconjs, 1 minute and 5 minutes res ectively. The sliles were firain d ani the reverse sides gently rinsed with ten mater. The counts obtained after theS' proce- dures are given in Table VII. Tne highest counts were obtained with the 1 minute immersion in Ceepryn-methylene blue. The plate counts here still consijeraoly hibner than these direct counts. an interesting observation was made on the above series of Smears. ”he extent of bacxground decolori- zation aypearei to be a function of the tine of immersion in the Ceeoryn-met-ylene blue stain. The longer periods of immersion resulted in greater decolorization. Immer- sion for 1 minute was eminently satisfactory as the backgrouni was colorless to pale blue, giving good con- trast with the stainei organisms. The "crest" of the film (16) retains the stain, permitting ease in focus— sing the microscope. The effect of defatting .he milk films before staining was assayed by examination of a parallel series of sliies. The films were pre,ared from a sample of 1.— skim milk immediately after seedinf with n. coli. One (‘ t of sliies was defatted 1 minute in xylene, rinsed U) «D in 95 percent ethanol for 1 minute and stained by the Breed technique and with the Ceegryn-methylene blue stain for jeriois of 1 ani 2 minutes. The duplicate -19- set was stained in identical manner, the defatting and rinsing procedures being omitted. The effect of de- fatting on the counts appeared to be nil. In one in— stance only was the count on non-defatted films higher than on defatted films Table IX). The observation was made that the bacteria were not concentrated in the fat droplets. It had been postulated that the organisms might be present in high cancentration in or on the sur- face of the fat globules and be lost in the defatting process. The possibility is not ruled out, as these bacteria might be inaccessible to staining and so be in- visible even when present in the milK film. In Table X, there is presented a final comparison of counts on all the organisms seeded in sterile sKim milk from 24 hour ejar slone cultures. Counts were made initially and after so minutes incubation at 5700. The agar plate counts and direct microscopic counts Where the smears have been stained by the Breed technique and the Ceenryn-methylene blue stain are compared. The di- rect counts using the Ceepryn-methylene blue stain are in 1:1 or 1:2 ratio to the plate counts initially and at co minutes, and are 5 to 60 times as great as the counts maie on samples stained with the breed stain. The application of the staining prOCedure developed in this investigation to the examination of raw milk 3 pplies was tested. Duplicate filns were prepared of 40 raw milk samples submitted to this laboratory for routine examination. One set was stained with Breed's stain and the other with the Ceepryn-methylene blue stain. Fourteen samples of the 4b examined(or o5 percent) gave significantly higher counts with the new stain proce- dure than with the Breed stain. at no time did the Ceepryn-methylene blue stain result in significantly lower counts than those obtained with tie Breed stain. The samples showing increased counts with the Ceepryn- methylene blue grocei re contained roi-form organisms which were not amparent in the duplicate films stained with the Breed stain. The data are presented in Table AI. In the course of these investigations, numerous other stAins and staining procedures were attempted. None gave satisfactory results. oyar's cell wall stai- ning procedure (lo) repeatedly dissolved the milk film from the slide. manwell's polychrome stain (Bu) was ‘ l) « 0 p H. {.3 excla' gly difficult to prepare, and when used at ph 7.4 dislocatei the milk film on the slide. attempts to counterstain the Ceepryn-methylene blue stain with dilute aqueous safranin proved futile. Instead of heightening the contrast between the stained organisms and the backgrouni, the contrast was diminished and the organisms were more difficult to locate. -Bl- DTQCUDOIQN Besides the interest in t‘e problem of coliform organisms failing to stain with the Breed stain, an- other problem of considerable interest to adjacent fields arises. This problem is the means by which the surface-active cation cetyl pyridinium chloride helps the absorption of methylene blue. In isso Moyer (28) showed that the electrokinetic potential of g. coll changed in various phases of the culture cycle. The electrophoretic\mobility of this organism was shown to drop immediately and by tne end of the lag phase to be at its lowest mobility, and, hence, lowest electrokinetic potential. The mobility remains low during the logarithmic period and than increases to its former high level during the period of negative logarithmic acceleration. The evidence presented in this investigation on the stain absorption bears a close relation to Moyer's observations. The ”physiologically old" cells seeded in the milk failed to absorb the basic methylene blue. These cells are shown by Moyer to have a high negative charge. Upon incubation, the charge drohs and the now "physioloEically young" cells absorb the stain. Dyar and Ordal (ll) showed that cetyl pyridinium chloride acts on E. coli to decrease the charge, reverse the charge and finally stabilize the charge at either a oositive or zero potential. This was demonstrated by them with 20 to 2B hour cultures. In effect then, -22- if the electrokinetic potential of the cell should be the determining factor in absorption and retention of a dye, we may have by treetment with the cationic agent produced th sane conlitions in "p ysiologicallv old" (D n "‘ '” ~~ ‘ s. ,, ..u ybioiologically Soune cells as are normally found in cells. The ”physiologically young" cells have been demonstrated to retain the stain in this study. The reason is not clear just why cells with a low negative or positive charge should stain with a dye like methylene blue, mhich itself is a cation. It would appear more likely that the more negztively charged cells shouli absorb the cationic dye, although the converse has been shown. The possmbility exists that methylene blue méy exist in a resonant structure with an iniuced negative pole, which would explain its absorption inder these conditions. No literature on this possibility is known to the author. The size of the cells in the course of reproduc- tion appears to have no bearing on the stain absorption. moyer has shown that the larger cells have greater perme- ability, but this appears to be a function of the reduced electrokinetic potential and not of the size per se. The phenomenon under consideration may be explained on the basis of the change in electrokinetic potential of the cells during their culture cycle. A method for -25- staining these recalcitrant organisms by changing the charge has been developed and a possible mechanism of action elucidated. It is interesting to speculate on the results of an investiéation of counts on raw milk samples by direct microscopic examination employing the Breed stain and the Ceepryn-methylene blue stain, the standard agar plate count and the measurement of the coliform index by a dilution technique. rerhaps the results of such a stuly would lead to the adoption of a rapid micro- SCOpic method for determining the numbers of coliform organisms in raw milk samples by a comparison of direct counts obtained using the Breed stain and the Ceepryn- methylene blue stain. The coliform organisms have been shown by uahlberg (9) to increase more rapidly than the rest of the flora in stored pasteurized milk. Because of the low storage temperatures it rould be expected that a preponderance of "physiolobically old" cells would be present. a technique, such as that SOUght for above, would be of immense value in the control of these stored milk sup lies. -24- SUMMARY 1. These studies have shown that only a minor portion of the coliform organisms seeded in milk from 24 hour agar slope cultures grown at JVOC are detected by the conventional direct microscopic examination. These findings are in agreement with the observations of Whitehead. 2. The hypothesis of a life cycle of coliform organ- isms as an explanation of the phenomenon is rejected. The hypothesis that coliform organisms fail to retain the stain under these conditions is upheld. o. A method for staining the milk films to bring the direct microscopic counts and the agar plate counts into closer correlation is developed and described. 4. The Ceepryn-methylene blue stain develOped was ap- plied to the routine examination of raw milK supplies. Of 40 samples stained in this manner, 05 percent were shown to yield significantly higher counts than with the Breed stain. 5. The theory of mode of staining is discussed in reference to the use of the cationic surface-active agent and the basic dye employed in the stain. W (i u I. ?aker, V. P. 1946 The direct nicrosc0pic count as a method for counting bacteria in pasteur- ized milk. J. Cairy 501., gs, 506. ?aker, ". P. and Peters, I. I. (Personal commun- ication). Wortree, A. L. (Unpublished data). Prannon, J. E. 1940 Studies of the resazurin test for milk. Tilk Plant Tonthly, gg, 51-55. ireed, ?. S. lEll The determination of the num- -‘ her of bacteria in rilk by eirect microsc0pical exanineticn. Centralhl. f. Ba t., II Abt., :9, 337-340. 3 v. 1"» P Drew, J. F. l ,9 The corparative accuracy of ‘/ C+ he direct nicrcsc0pic and agar plate retnoos F1 :3 deterrinin; nurbers of bacteria in milk. J. Fairy Sci., lg, 304-319. Drew, J. D. anl fireed, R. S. 1945 The bacteria ”count"- an estirate capable of accurate inter- \ Uretation. A. Jo Po Po, i, 5E.\,)3"’1"880 ryan, C. 8., Vallwann, Y. L. and Turney, G. J. l§44 Sore microsc0pic technics for determining ' 1 oactericloalcal quality of milk. “icn. Agr. Dahller: A. C. 1946 The relationship of the growth of all bacteria and coliforn bacteria in pasteurizefi milk held at re i? atures. J. Dairy Sci., 29, {bl-‘55. lO. Cyar, T. T. 1947 A cell wall stain employing a cationic surface-active agent as a noroant. J. (D ?act., :1, 4 \0 ll. Pyar, ’. T. and Crdal, E. J. 1946 Electrokinet— ic stuoies on bacterial surfaces. I. the effects of surface-active agents on the elect- rophoretic wobilities of bacteria. J. ?act., El, 149-157. 12. Erb, E. T. 1929 A rapid stain for the direct niCTOSCOpiC examination of milk. J. Lab. and Clin. T"ed” l3, 377-379. I}. Banter, 3. T. 1938 Fairy bacteriology. 2nd ed., John Wiley and Sons, Inc., :ew Eork, 1-19. 14. Jenkins, H. 1941 A conparison of nicroscopic and 3d C T”? agar plate counts on raw milk. J. Vilk Tech., 3, 514-517. H Levowitz, T. l€44 Reproducible data by micro- scopic rethod. Assoc. ”Ull- INE- ASSOC- 311K \4‘ fealers, 36th year, lE;-l“0. 15. Tallvann, W. L. and eran, O. S. 1943 Emerg- ency use of the laboratory during the war. ASSOC 0 7i111.]. o I '3 .t. Assoc. Tilk Sealers, 32th r“ {I ;‘~ I“, year, :J-lic. 17. 20. 21. T"allnann, W.L., Pryan, 3.8. an5 Saten, W.:. 1944 ‘ O. A stuay of nations for the examination of raw at: nil? zith suggested irproverents. J. ”ilk Tech., Z, Eli-3&1. 1' fallnann, T. L., Bryan, C. S. and Fox, W. ’L‘ 1941 A new microsc0pic procedure for the be- 4. b (D ctin r“? ‘ln p and locating of the source of thermo- duric organisms in nilk. J. ilk Tech., 4, 195-159. Tallnann, W. L. and Churchill, I. S. 1942 A rapid test for Geterainin; bacteria in liQUid azine, 43,40)- eaaS. C. S. 75; and Poultry 'a V \v' 407 427—429. \ 'anwell, R. P. 1945 The J. S. B. stain for «D (.0 c4 r4 5:0 J to :3 Q; C) H H :5 (D C) W 0 blood parasit lO7?-1032. Foyer, L. S. 1936 Changes in the electrohinetic potential of bacteria at different phases of the culture cycle. J. ?act., 23, 453-464. 'udxe, C. S. 1§30 A life cycle of a thermophil- 7“ ic orfanisr. Proc. Soc. 3X9. ‘iol. and Ted., R) 22) 262‘ 03. Robertson, A. E. 1921 The relation between bact- erial counts from milk as obtained by nicro- SCOpic and plate counts. 5. E. State EXp. Sta., Tech. Pull. 86. 24. Shutt, D. P. 1947 A rapid staining nethod for the microSCOpical examination of milk. Stain PO \ n .. Standard rethods for the examination of dairy products. 3th e€., A. P. H. A., 1941. ‘r‘ k Patrons, G. H. Jr. and Doan, n. J. 1947 Cbser- [U r“\\ vations on the use of a nodified direct micro- SCOpic retbod for estimating bacterial quality of raw and pasteurized nilk. J. 111K ans ECod m Tech., 129, 251-9-271. 27. Thitehead, E. R. 1931 A note on the direct nicrosc0pic count of bacteria in milk. J. Dairy Res., a, :3. Po STAIN FOnMJLAga Modified Breed stain: 1o ml saturated alcoholic methylene blue (certified for bacteriolO¢ical use). 90 ml 50 percent ethanol. Mallmann's acid stain: 0.2 gm methylene blue (certified for bacterioloéical use) 100 ml 95 percent ethanol 1 m1 concentrated hydrochloric acid Ceepryn-methylene blue stain: 5.4 ml 10 percent aqueous cetyl pyridinium chloride 10 ml saturated alcoholic methylene blue (certified for bacteriolobical use) bc.b m1 at percent ethanol * All staining was done in Goblin staining jars except where otherwise indicated. ii ELUOJIML‘J NULL-U bulinlN IN G I’HQGILULJ Ind Air dry sliaes in horizontal position. Fix by heating in flame until too hot to hold with fingers. Allow to cool. befat 1 minute in xylene. urxin slide of xylene. Rinse 1 minute in asp ethanol. brain slide of ethanol. Inmerse sliie in Ceepryn—methylene blue solution. Urain excess stain from slide. flashreverse siie of the slide in a gentle stream of tap water. Drain and dry. iii TAaLn I Comparison of plate ans direct counts of bacteria :3 4 y...» H i a“ Q) H) C+ (D *‘S u" ,eriods of incubation at 3700. bacterial counts*(thousands) Incubated (rinutes) O 30 60 90 120 Crganiss :1 921:. plate 29g 330 327 410 540 direct 9 42 6O 162 35 g. aerogenes # plate 65 legO 1310 1290 1750 direct u 06 120 795 1250 Intermediate ‘ ‘ _ , plate 252 2:5 565 1190 lCcO direct 18 96 2(1 10c0 lCBO :. caseolyticus plate 00 720 620 -~ 690 direct 510 510 480 -- 600 * Average of duplicate counts. iv TAVLE II Comparison bf plate counts and direct counts of bacteria seeded in sterile skim nilk. 2acterial counts (thousands) rfi* "151t161 ~0_0 fter 60 mins. eranisn Plate Cirect Plate Direct é- aero:enes 375 43 550 515 653 48 1310 120 1350 12 -- -- 1000 36 1340 900 3. 0011 4230 102 6000 6060 5 y 299 0 327 67 ; 630 50 -- -- g 1470 24 1560 450 j i 370 54 -- -- i ! 1800 42 —- -- 5 ‘ Id :Interrediate 5450 165 7350 r400 1 p 252 18 565 261 1 5 448 36 810 210 5 g 3250 192 21900 3450 i 1970 12 -- -— l 7000 72 7600 5000 i E. caseolyt- 160 192 205 348 Laue 1020 500 —— —- § 700 510 620 480 itially and with Coroarison of cultures fr0r after *3 J;- .0 L“ Li) III difterent sources . direct counts of nilk samples, 60 rinutes incubation, w" in- seeded 99 .1 bacterial counts (thousands) 4; fi“ 11L seeded with Initial After 60 minutes 6 hour broth 2. 0011 48 50 Q. aerogenes 78 90 Intermwmiiate 150 396 48 hour agar lepe 11.0011 6 36 a. eero ease 15 108 Internedirte 4 132 24 hour agar slprWI—“E—H“ E. goli 3 72 a. aerogenes 9 495 Internediate 4 162 (D *5 D (N .19 of duplicate counts. saaples, when seeded vi EHflEIV Comparison of direct and plate counts of milk ivitti cultures initially and after 60 minutes incubation, from different sources. '3" . . Pacterial counts (tnousands) hilk seeded with Initial After 60 minutes Plate} Direct Plate Direct 5 hour broth g. 001; 156 144 152 246 g. aerogeres 26 30 35 24 IJterrediate*% 43 150 160 252 4 69 198 510 432 24 hour brotn Q g. 9911 \ 145 73 160 156 6- aerogenes i 110 12 226 216 Interrediate** r 77 135 38 312 712 263 1220 1170 24 hour agar slope E. coli 279 9 327 60 g. aeroyeres 653 48 I310 120 Interxediate 262 18 565 261 Average of duplicate counts. See text. vii TA“LE V Comparison of direct microsc0pic counts when replicate sli as are stained with Ereed's stain and lallnann‘s acid stain after treatment with surface- active agents. . - . ‘3? - :1 ect oacter1a1 counts (thousands) r g. 991; Internediate e“ Surface-active agent Ere d Tallnann Treed tallmann nil 6 7.5 6 7.5 alkyl dimethyl benZyl annoniun chloride 16 ** 15 && cetyl nyridiniuw chloride 9 2* 53 %% Tergitol 4 24 %%* 36 %%* Tergitcfl.'7 red: 22* *%w *%x DUponal L-141 66 l“ 51 4 * Average of quadrUplicete counts. “* Patchy staining, abundance of detritus and crystals. *** Milk film washed off slide. TA?LE VI Comoarison of direct ricroscOpic counts when the silk files have been stained with the Vreed stain and the Ceenryn-rethylene blue stain containing varied concentrations of cetyl pyridinium chloride. The agar plate counts are included for corparison. Pacterial counts (thousands) Crganism and stain Initial After 60 minutes enployed* Eirect Plate Direct Plate : 9.91.1. 2reed 24 1470 450 1560 Ceeoryn-“.2. 9/100 132 1140 Ceeoryn-".2. 1/200 60 480 Ceepryn-T.?. 1/50 12 660 T. ceseolyticus _9P8¢¢ 510 700 480 62c Ceepryn-V.?. T/100 450 '80 Ceepryn-“.2. 7/200 420 500 Ceepryn-T.2. T/SO 180 72 * Ceenryn-T.2. is t?e Ceepryn-rethylene blue stain in 50 percent ethanol, with the indic7ted con entrat- ion of cetyl pyridinium chloride. > Conparis on of one d2 y old and thirty day old 2P8f3T9t1018 of the Ceepryn-nethylene blue stain, the Ceepryn-rethvlene blue stain prepared with 70 percent ethanol, the reed stain and the standard plate count. 2acteria1 counts (thousands Organisr and stain Initial After 60 sins erployedt Direct Plate Direct Plate 1. aerogeues 1000 1340 “reed 36 900 Ceepryn-‘.V. 30 1 180 1260 Ceenryn- .2. 1 day 160 1120 Ceeo. ryn-T.?. Op ethanol 48 1050 Interredlate 7000 7600 ”reed 72 30 CO eepryn-V.B. 30 day 640 5400 Ceepryn-T.3. 1 day 540 5800 Ceepryn-V.B. 70% ethanol 72 5200 * Ceepryn-T.B. is the Ceepryn-tethylene blue stain containing N/100 cetyl pyridinium chloride in 30 per V cent etne n01 exceot where indicated ethanol. as 70 percent '1 1 Corparison of counts obtained or direct xicro- sCOpic exarination when the time of inmersion in *4. U) Ceeeryn— ethylene blue sta n i varied, when stained with r3reed's stain and by the m tandard plate count. Wacterial counts*(thousands) Crganis: and nethod 3- go i Intermed— g. aerc 3. 033304 1111212. i c of examination** i iate :eneg ., ‘ l girect, stained oy; 7 ‘reed 30 12 12 500 Ceepryn-T.R. 30 sec 42 2 30 SSO Ceepryn-T.B. 1 win 138 100 210 525 Ceepryn-".W. 5 min 84 E6 170 520 Plate count 630 10 D «l O H \N K 3‘: O N H m * Average of duplicate counts made irmeaiately after inoculation. I \v 9? 'l\‘ F3 pqa :e Ceepryn-rethylene blue stain is the solution containing Y/100 cetyl pyridiniun chloride in 30 per- cent ethanol. xi TA2LE IX Corpariron of the standard plate counts and the direct nicrcsconic counts obtained on milk filas de- fatted and not defatted Tefore staining by different :ethcds. 411 silk satples have been seeded with E. coli. 2acterial counts*(thousands) m o _ o a 1 11re incuaatec and n . ? ,+a D r 7 = 1 - = rethod exanined by: piles eefatceu 1ilms not uefatten Initial counts Tirect 2reed 54 30 Ceebryn-T.?. l nin 150 50 Ceepryn-T.2. 2 sin 50 140 Plate 370 l 370 After 90 minutes 1 Direct Wreed _ 2’5 500 Ceepryn-T.2. l Fifi ; 320 l 370 Ceepryn-T.V. 2 nin I 515 ; 375 I Plate 380 l 380 * Average of duelicate counts. Comparison of a xii gar plate counts and direct counts emnlovinq the fireed and Ceeor n-reth71ene blue stains. . ., 1- 1 J 1 a Initial After 60 rirute .. a q 1 1. ‘ 1 1, liitheehe“ 2reedrggepryn-T? Plate Lreedrgggpryn ;3 Plate 3' 0011 102 3130 4230 6050 5500 6000 Internediate 192 4770 C250 9450 11100 21500 1. aero:enes 48 204 375 :15 615 550 :. caseclgticus 152 174 1(0 543 270 205 V W Average of duplicate counts. xiii :A“L? x1 Coaparison of direct aicrosco;ic counts of raw using the 2reed stain and the Ceepryn- rethylene blue stain on duplicate snears of each Direct 2acteria1 counts (thousands) Counts increased using Counts in agreenent us- Ceeeryn-T.s. stain in; either stain E Rreed Ceepryn-T.P. # “reed Ceepryn—X.7. 1 12 84 2 12 24 5 90 750 3 12 48 7 12 660 4 120 156 ll 50 156 6 12 12 15 91 258 8 12 24 18 315 600 9 90 E4 21 540 950 10 500 170 23 360 900 12 12 3: 24 400 T50 13 200 238 25 60 240 14 50 36 28 300 1200 16 700 800 29 oo 900 17 50 5 32 90 230 19 75 S4 33 305 1100 20 100 144 22 1500 1500 6 2000 2700 27 520 650 30 120 150 31 6‘ 6 34 60 E 35 120 120 36 1500 1300 37 180 190 8 75 84 39 25 48 40 80 72 ”11111” 1111111111111?“ 1111 11111111 “ 281327