: / 19x79

GYPOGEN Zils OF DIPLOID AND HAPLOID CULTURES DERIVED FROM
BACTERIUM SOLI, SiRAIN K-12

Ethelyn Lively

The isolation of diploid cultures of Bacteriua coll, strain K-12, (Lederberg, 1949 )
has provided an avenue of approach to cytogenetics of bacteria. In this work, "diploida®™
refers to unstable prototrophs which arose from crosses of auxotrophic mutants, and
which are heterozygous for one or more sugar fermentation loci, in wnich the parents
differ. They are identified geneticglfy by plating on a complete nutrient agar medium
containing eosin-methylene blue indicator, plus the ap»ropriate sugar. If lactose is
added, for instance, this medium is called EMB-Lac. Since the ability to ferment sugar
is dominant .o its absence, a diploid nékerozygous for lactose fermentation wiitraerk,
lac #co-oniea on this medium, but ligt sectors will appear as the Lac - *aploid conponent
segregates during the growth of the colony. Such a strain is called Lac v (variegated
colonies). The following laboratory abbreviations for media will be waed in this paper:

=MB.....-Eosin- methylene blue, complete (peptone)

EMS.....Hosin- metrylene blue, synthetic, which does not support the grouth of

auxotrophs.
. NSA.....Nutrient saiine agar, Difco plus .5% NaCl.
he current inves:igation deals with a cytological Any éds8igadtigx comparison of

divloids with their Wreloid parents, segregants, and wild type K-12. The methods employed

Ph
“ay

are nuclear staining by Robinow's technique (Klieneberger-iobel, 1950), and more recently,
observation of living celis with a dark phase contrast microscope. Diploid cultures have seer
been distincuished from naploid by these mechods but it has not been conclusively shown
that the difference desends on the number or size of the chromatinic sbructures which
nobinow has called o:-romosomes (Dubo Sep 1945).
For nuclear staining, blocks of agar from a plate spread wibh bacteria are cut
ovt and nlaced in small “etri dishes Por incubation. At desired time intervals, dishes
are re oved from ibe in ubator and inverte’ ove small wide-mouthed bottles of Osmic a-:id.
-"e vapor fixes the cells as they grow on “ns surface of agar; then, an impgzesisn of she
grow is printed on a coverslip. “he bacteria on th: coverslio are post fixed in

~

Sehaudinn(’s reagent, washed in 7045 alsoval, water, cold normal HOl, and hydrolysed for

3
Patt

 
y

10 minutes in normal Hl at 60° Se. °-ey are returned to cold Hl, water, phosphate buffer
as DHT, stained with itiemsa for 30 minutes, rinsed in buffer and mounted onfa slide in
the water soluble resin, Abopon. \/

Observations o° living cells can ;e made on similar agar sevtions cut from the
same plate, provided a tnin plate of colorless medium (NSA) is used. [be small section
3g mounted between slide and Coverslip which have teen sterilized by flaming, and the
edges are sealed with manometer grease. \/ Then che slide can be incubated between
observations or xept at room temperature on the microscope stage. This method for phase
microscopy ig based on that described by Stempen(1950), working with B.coli and Proteus
vulgaris. He identified light bands in living cells with the chromatinic structures that
stain with Feulzen and siemsa. These light bands are visable in the series of pictures
of K-12 (1--5). In this experiment, the lag phase was 3Zhours, and division time thereafter,
about 45 minutes. No differentiation at all is visable in the earliest picture made before
division began. At thés stage, cells O¢/Livthé on stained slides are also small and
poorly differentiated. (see piztures 7--9).

The same Bausch £ Lomb research microscope with pase contrast aedessories is used
for vnase and for bright field photography. For the former, illumination is provided by
a carbon arc lamp; for the latter, critical illumination from a ribbon filament lamp is
used, with a Wratten 8B (green) filter. Tne camera is a Bausch & Lomb L type.

In pictures 21, 22, 23, living cells under phase contrast are compared with stained
preparations from the same culture, photograpbed with the same 97X phase ob’ ective, and
in bright field, with a 90X apochromatic ob /ective. ‘The phase pictures show that osmic
acid fixation has not shrunken the cells appreciably. All these pi=tures are contact

prints from 5 ¥7 Panatomic~X or Suver- Panciro-press Easimen Kodak film. Differences

 

1. Staining reagen:s:
Scv-audi-n's reagent: Hgvl, lOgms.jHo0, 2002¢; Ethyl alcohol, 100 <c.
Giemsa stain: National Aniline Div. Allied hem. and Dye -vorp.
Dilute, about 5 drops in 10 :c Hod.
Abopon: Glyco Products Jo. Dilute with bea.ing, 2 prts to lprt. Hod).
2. Equal parts lanolan and Vasolin.

 
3

in nagnif icationwere obtained by using @ifferent microscope lenses, or varying the extén-
sion of the camera bellows.

Different stained preparations of th: same strain show a rather wide range of vari-~
ability in size of cells and appearance of cshromatinic structures. Uften the source of
variation is unknown and a particular aspect is not always reproducable. In general,
the appearance of hapioid K-12, during the logarithmig growth phase agrees with the
earlier similar study of B. coli by Robinow. (Dubose, 1945).See pictures 7--17 and 24-29
for culture cycle series of K-12 on NBA at 23° and 37°C. sote the symetry of adjacent
chromatinic s:ru tures that obviously came from a recent division perpendicular to she
long ax's of the Cell. Ip,picture 27, for example, a number of cells contain two sets of
double suruciures that look like anap’ase figures. Hach is symetrical with respect to the
other set and with sespect to its two halves. Few cells are seen with fess than two
distinct rods, but this ‘s probably a matter of inability to resolve the very young cells
(pictures 7, 24,25). fois interpretation is the sams as Robinow's. If each rod isa
curomosome, as he b:lieves, the symetry of the paired structures indicates that they are
chromatids, rat-er than two separate chromosomes, and the nuzlear unit is probably one
chromosome wich may divide severai t'mes prior to :ell division. This agrees with the
genetic evidence for one linkage group in K-12 (Lederberg, 1947).

4s cells increase in size, permitting better resoluvion, of the chromatinic
structures, they tend to change from 2 :indensed rods to th nner, more numeroud bodies.

( compare pi ivre 27 with 28 and with 30) The same tendency is sometimes noted in K-12
grown at room tomperaiure as compared with 37 v. (-ompare 10 wits 27). Sometimes haplo‘d
cultures in this stage, approach tie appearance of diploid -:uliures where the’ occurtayce of
larger celis with relativééy disperse chromatinic structures is mu-h more regular.

Since diploids are conviinvally segrega:ing, impressions from plates of complete
media give mixtures on tre slides of various proportions of diplo’d and naploid cells.
Yiplo'dy of She inoculum must be verified by heterozygosity tests for the fermentation of
some sugar. Of course, there is no meitcd of characterizing single cells from a Pix ed
prepara’.on. It is not proven that tie large cells with relatively disperse chromatin
are diploid, bus watever is rosponsible for their occurfnce is more affective in dipioid

tean in haol id cultures.

 
7

fo test an ooserver's ability to differentiate between haploids and diploids
under idevsical sonditions, some experiments were desi»ned as follows: A Lactose het-
erozygote was plated on EVS- Lac. Instead of variegated colonies,here, a lac v cell
will usva ly produce a pure Lac f/colony secause the medium lacks the amino acids neces-

ee
sary for the growt: of most of tne segrerants. AP ew prototroshnic segregants will occur

on te same plate, and trose which are Lac # cannst be distinguished foom the diploid
colonies except by restrea “ing on =MB-Lac. Using suspensions of single Lac fcolonies
as inocula, it has been possible ‘to prediet which will be diploid Prom the cytological
appearance of the bacteria on the EMB plate after 3 - 5 hours of growth at 37°59, and
before the genetic evidence became available the next day.

Because of reporss fr-m other laboratories on staining bacteria infected with
ba:teriophage, (Luria,1950), it was thought that the presence in K-12 of the lysogenic
phage , Lambda, might be affectang the apsearance of c-romatin in both diploids and
nacoloi?s. ‘me experiment shows that this is probaébly not so. Dr. Esther Lederberg
provided the stocks from which both lysogenic and )-Pree diploids sould be syntresized
for purposes of couparison. (seef pictures 3C -- 35) One parent, W-5&8 is like K-12
in that it is lysogeni:, resistant to the phage it ca-ries. The other, W-1248 is non-
lysogenic and resistant to d - It was derived from a sensitive strain, W-518 by selec-
tion with puage. VSytologically, there is no consistent difference between the lysogenic
and non-lysogenic naoloids. A cross was made by the EMS plating technique, and Lacv
solonies were selected. (Lederberg, 1949) In cross streak tests with the sensitive sirain,

caused
one diploid d/pfigd no lysis. <= is d -free diploid, H-232, is cytologically indistingvis’-
able from those previously examined and from a A #d ploid isolated from the same cross.
Apparently, X cannot be detected by t:.is staining metiod.

The interpretation toat granular chromatinic structures, characteristic of diploid
cultures, represent two homologous ‘hromosomes, compared to the one in haploids, is an
inference Crom genética, which this cysologizal evidence neither supports nor disproves.
“here does nut anpear to me twich as much c romatin in the large cells as in those from
aploid cultures, sus Dr. Hans Ris his sugzested an analogy to ihe spermatocygés in insects
of the X #0 sex type, where te unpaired X may be heteropysnotic and appear at meiotic

metaphase, “ust as lense as the other -hromosomes w ith are paired.

 
Ss

“wo toloid stocks have proved »varti ularly useful for cytological experiments
because the characteristic chromatinic structure is very pronounced. [ ey are H-226
( Lac 1/4 v Gd v Mal v Mtl vXylv) and H-267 ( same as H-226, but also segregating for

~ 7 Cpr nee SF- 1, 12-45) .

streptomy:in resistan:se). This heterozyzosify for so many characters, (including maliose
ferm-ntation, which is usually hemizygous) may mean that they actually have more comp ete
nuclei, than the “aberrant heterozygotes" previously examined. (Lederberg 1949). They
are relativély stable diploids. Vaintained in liquid minimal medium , a high percentage
of the cells remain diploid and their segregation can be observed when they are plated on
Ev3 or NSA. ‘o make the impression siided used for photographs 46- 49, about 10? cells
were spread on an #M3B p ate and fixed after about 5 vours growth. ‘wo cell sizes are
very d s.inct. When the micro-colonies seen at low magnification (46) are resolved, £
a 48, 39), some are seen to consis’ of uniformly short cells with 2 or 4 compact nuclear
bodies. Otrers consist of bigger and much onger cells which often hve their chromatin
neatly distributed in aggregates of smali granules. Some microcolonies contain both types
of sell in sectors. the obvious assumption is tat the short cells are haploid, the

tong ones diploid, and the mixed microcolonies arose from diploid cells that segregated.

Agadmn, there is no direct proof of these identities, but the folloging lines of
evidence are now being followed with the aim of describing the cytology of diploide and
naploids ona cellular, rather shan a cultural basis:

1 - vomparison of cu tures from gen@ticly known segregants and diploids:

Practicaily; all the segregants from :-226 and H-267 are Lac - because the diplsi ds
are seterozygous Pro tne two -losely linked Lacy and Lacy loci. All lac - colonies on an
ENB plate are segreganis and Lhe two narental Lac - loci are distinguishable as slightly
differen: s7ades of lignt colonies. A few cultures were prepared from H-226 sezreganis
identified in thés way and were found to consist of uniformly smort ce!ls. (Pistures35-41)
It has not been determined whether they are consistently smaller than the parent strains,

as the suall ceils in she mixed clones Prom dioloids seem to 3e.

XN

2 - Identifying chara-teristic staining types with specific sizes o” living cells, and

subsequeni.ly 2 ara Lerizing the !!ving ce.ls sy observing their “abit and rate of

grow.h or by gene:ic neans, or bo-h.

 
£

3- Sonparison of series of s:ained preparations of unese diploids under conditions
known to produce abnormally sigh proportions of haploid cells. Methods 2 and 3 will

$. 1

be discussed ‘together

Grow:h rate experiments comparing -aploids and diol ids by usual culture plating
technicues hare now been autenpted because of the difficulty of maintaining diploid cultures
owever, d‘rect information on growth rates comes from Zelle's (1951) experiments with
these sura’ns. He separated single cells with a micro- manipulator, watcored them grow into
micro-zolon.es and tven picked Shem vp and identified tnem cenébtacaly. Diploid pedigrees
siow that wen segregation oc.urs, one cell divides to form one which is vaploid and
one, still diploid. ‘e has observed shat the »nanloid grows faster.
In my own experience, Dark phase contrast observations of living bacteria have
poen telyfvl in estabdlis ing growt” rates and obserfving inter and intra clonal size
variation, but ha e een little -elp, so far, in clarifying the navure of the chromatinic
structures. In different :xperimenis, the lag phase and division time of the same strain
at room temperature, h ve not always been the same. Mo effort was made to keep the
bemperaiLure constant on «ne microscope stage and other sources of variation between
experinents are the ‘oncentration o. bacteria and the thickness and moisture content of
the awar. “he series of K-12 (1--6) at 27 °@ is probably rovg»ly comparable in hours
so she one of H-267 (S51 --69) for w.ich che plates wer: incubated at 47°S for the first
45 -sinut es, and subsequently grown at room temperature, which was 23°C on that day.
It is obvious that division proceeded faster in che K-12 series. in one experiment
che growth rate and general living aopearance of the parents of H-267, were found to be
similar to ¥-12. (pi-tures 5C -50}.
In the diploid series (61 - 69) there are two distinc’ cell sizes “very small"
and "medium". The growth of four sells can be traced separately for 5} hvars. Jells
5 and 4 were apparently dead when plat d. No. 3 produced uniformly emall cells from the
beginning and divided at a consistently faster fate tian the other three. Nos. 1,2, and
4 produces eslls of a uniform medi'm size excep. for one "smake" in clone 1. It seemsrea-
sonable to assume that no 43 was naploid and the others diploid al time of plating.

vue glide was left at room te perature over n'gnt and the same field obser ed the next

mornigg (69). [here was no obvious change in the propprtion of very ssall to medium sells,
7

and this field seemed .o be representative of the whole slide. After pic:ure 69 was
made, she cells were scraped from the agar and streaxed on EMB—- Lac. The result was
vos:ly Lac- colonies, ‘ndicating al..ost vomplete segregation. “nis seems to contradict
the assumption chat all the med ium cells in pi-ture 49 are viable diploids. It is
possibie thal many of the celis were dead. The slide was no. watched long enough on
the second day jo determ‘ne whether division was still taking place. Pernaps the high¢
proportion of “aploid secreganis was caused vy differential survival rather than complete
sefregation. Vorresponding stained slides of the 21 hr. plate might have been informative,
but they were not made because slides from such crowded plates usually show very litile
or are difficivlt so interpret. Picture 29 was made from such a preparation of K-12. If
Lhe large number of ghosts represent dead cells, and thse oaly 1 ving ones are those

has
containing chromatin, then it is obvious that such a ppopulation gbfAtyGy? presented a
wide opportunity for selective action. Possibly this is the way in wich acid production
by growing bacteria actsto increase the proportion of haploids. This is a known effect
of artificially lowering the pH of cu-ture medium (Lederberg, unpublished). If this idea
is correct, one mizht expect stained slides of a preparation such as pi.:ure 69 to show
patches of short ve lls containing chromatin, corresponding to the center clone in the picture.

In comparing stained with living cells certain technical d/fferences must be kept
in mind. For fixing and staining, it is usually prac.ical Lo plate a lower dilution of
a culture tran can be ueed for observing growth. Even sections from the same plate are
sub’ect so different conditions under the microscope. 3esides periodic exposure to a
source of ligrt and heat, shey differ in that tiey are growing anaerobicadly in contact with
a cover slip. It is possible to cover the se tions wich will be used Por fixation ina
similar manner and fix from the -overslip rather than the agar. A very listie experi enta-
tion wit. these alternate metcods indicated thai there is little difference, but the agar

gs?
method has been used here because it szemed to result in bet er stains, and,the question
of variation still exists.

In spite of “inhes¢t limitations, similarities setween stained and living cells of the
same age are close enough to carse little hesitation in identifying the medium cells (in
aa en a mee ro
icture 55) with the selis cantaining "diploid{titelei" that make up the majority of the

population of pictures 73 -75, and the fewer very smail cells {nm 73 -75) probably represent
&

the type in -lane 3 of the living series.

‘sis series was cun as a control in parallel with anotheplasing from the same
culture after Mltra -violet treatment. A 1/10 dilution of the culture was irradiated for
20 minutes at 5Ccns. Immediate olating from Petated and control on EMB - Lac showed
that survival (in the dark) was about 40%. The preparation which was observed after

45 minutes, and photographed at intervals

4

wievreafter may have been sub ect to light
reactivation (Kelner, 1949) from the arc lamp. [re control was observed on a separate
vlock of azarpn the sare slide, and photograp s were taken alternately. (61 -91).

One effect of Ultra-violet treatment of diploids is a greatly increased proportion
of sezregants. Here, the assay plates showed 6C% more Lac - colonies after treatment.
however, it is dificult to count these accuratsly because many colonies that appear neg-
ative after 24 hours develop tiny Lac fcenters when incubated for another night, indi-
cating tne delayed growth of one or a few diploid cella.

In the U.V¥. series (77 -91) again assuming that the very small cells a-re raploid,

~ Th Sine di Alest2g amt ben actend-
there is evidence for true segremat‘on during the first few houss.7 In pictures 83-85,
Lhere are perhaps some of she medium sized cells in clones 1 and 4, but most of the pop-
ulation consists of the very srall and a c:ird "very large" type. One of these was produced
at the first dividion of ceilNo.6, while its sister eell produced many of the small type.
No.2 never produced anyt ing but very small cells, like iio3 in the control. Nos. 3,and 5
probably consist entirely of the very large , slow growing type, although t is is hard to
determine because bot clones have merged witi neig:bors. ‘ie very large cells also occur
in the zontrol, but less frequently (pictures 70 - 72) This "type" is probably a heter-
ogenous result of many iifferent effects, but the tentative suggestion is offered that
some of them are cells wh ch are capable of giving rise to diploids at a later time, thus
accountéang for the delayed aspearance of diploid centers in haploid colonies.

_ “ne shree sizes are also clearly distingu shable in stained slides of the treated
preparation, (16 - 23) but the typical “diploid apsearance" is conspicuousl,; rare.: Note
(in 21 - 23) that the very large cells contaixing disperss chromatin appear mo sty dark
in phase contrast, as do the very large living eells. his is probably due to their thick=

ness ratoer than their interna! structure. he type containing condensed c romatin (black
in 22, or brigit spots, in 23) are not so essily related to any of the living cells at

a comparable time. They may correspond to che snake with the clear area in the center

in picture 91. After 21 hours such clear spots had occured in practically all the very
large celis. Tey showed no furtter change when observed an hour late. However these
light spots and square areas in living cells may represent some phenomenon entirely differs
ent from the condensation of ¢c>romatin.

In a previous irradiation experiment with H-226 (no photographs) many snakes
Similar .o the one in picture 91 were observed after 5 hours at room temperature. One
which was wa ced for two hours was alive, as evidenced by two unequal divisions giving
rise to two small cells from one end. (similar to pisture G4, clone3) T here were variou 8
‘brignt bands and spots in this snake, that persisted for a wile, then grew smaller and
after the first hour, disapoeared completely, leaving the smake hoxozsneously dark. The
firs. division occured imsediately after their disapoearance. In another half hour she ape
spots reappeared in the same areas, chen disappeared, and the second division occured.

In the same culture, a number of cells having persisvent J7Zgf/oright bands, showed no
signs of life.

Ih is not clear how the light area arose in pisture 91. ‘Its position correspends
to a spot in the previous picture tnat looks like a constriction, but may be the beginning
of the lignt area. Another, though less spectacular example, occurs in the unsreated
saploid snake in sictures 53 - 55. Apparent constrictions of she right end (Picture53)
have become prominan: ligh: spots (in 54) and disapseared comp)stely (in 55) No inter-
pretation will be attempted wentil this p).snomenon has been reproduced and amd more
carefully s:udied.

Ano he-afPeature of tbe H-226 experiment fas the lysis of many of the cells, pre-
sumably due to U.V. activation and liberation of l-sogenic phage. (iwoff and Deibr::ck,
unpublisied) roma field of six :e.1ls, under observation, turee disapoeared during the
Pout hour afier treatment, leaving only Paint ghosts. ‘his was observed only »>nee in
the H® 267 experimani: ino - 91 the long cell in che corner of pisture 90 (oui of

fosus) was not Shere she noxt worning.
10

. SUM mM A RY :

Diploi: cultures of 3. ¢ ‘i, strdin K-12 have been dissinguisned from haploid
culcures by means of nuclear staining. The former have a high proportion of relatively
large cells with a distinctive type of chromatin'c s.uructure.

“he presence or absence of the lysogenic phage , lambda probably does not affedt
the gph édxaAeE/d¥/éi dy staining reaction of eit=er diploids or haploids.

In certain relative y stable diploid stocks, known to be heverozygous for a
large number of factors, two ce 1 sizes are particularly distinct and can be shown to
be localized in microcs!onies, suggesting clonal growth from large diploid and smail,
segregant ceils.

Preliminary studies of thse s ocxs by observation of living cells with a
dark vhase tantrast migrossope “end ‘so svoport this “ypothesis, and comparison of  ,:
living and stained cells from the same culture indicated that the tipe of chromat‘nic
structure characteristic of dipioid cultures, occurs predominaktly in diploid cells.

“lira-violet irradiation of dinloids causes some haploidization, for which

trere is parallel c tological and pen@tic evidence. Irradiation also has a specific

Ph

erf

oO

st on she growth navit and the nuclear material of some ceils. Ts may be correlated

with a delayed srowth of diploid :3lls.

References:

Dubose, R. J. The Bacter‘al dell; Addendum by 3. ¥. R2obinow, Harvard, (1945)
Kelner, A. Proc. Nat. Acad. Sci. 35, 73 (1949) |

Slieneberzer-Nobel, E. Quart. J. ‘fier. Sci. 91, 340 (1950)

Lederberg,J. Genetics 32, 3505 (1947)

Lederberg, J. Proc. Nat. Acad. Sci. 35, 178 (1949)

luria, S. E. Scienzellll, no.2889,507 (1950)

Steusen, H. J. Bact. 60, 81 (1950)

Zelle, M. R. 4 Lederberg, J. J. Bact. 61, 351 (1951)