J. Cell Set. 34, 279-288 (1978) 279Printed in Grea t Britain © Company of Biologists Limited igy8

RNA TRANSCRIPTION ON THE GIANT

LATERAL LOOPS OF THE LAMPBRUSH

CHROMOSOMES OF THE AMERICAN NEWT

NOTOPHTHALMUS VIRIDESCENS

S. E. HARTLEY* AND H. G. CALLANDepartment of Zoology, University of St Andretos,Fife, KY16 9TS, Scotland

SUMMARYBy counting silver grains in autoradiographs of lampbrush chromosomes the rates of

incorporation of [3H]adenine, pHJcytidine, [*H]guanosine and ["HJuridine, administeredseparately, into RNA on the giant loops of chromosome II of Notophtlialmus viridescens, werecompared with the rates of incorporation of these same precursors into RNA on other, un-identified loops. The overall rate of RNA transcription on the giant loops is only about halfthat on the generality of other loops, and the RNA transcribed on the giant loops is computedto have a base ratio of approximately 25 A:39 C:o, G:27 U, implying that there must beabout 4 times as many guanine residues on the transcribed as on the non-transcribed strandof the giant loops' DNA.

INTRODUCTION

The evidence for transcription of RNA on the lateral loops of amphibian lampbrushchromosomes was established by the observations of several early workers (Gall,1958, 1963; Gall & Callan, 1962; Izawa, Allfrey & Mirsky, 1963). The overwhelmingmajority of lampbrush loops are found to be uniformly labelled within an hour ortwo following the provision of radioactive RNA precursors to oocytes. This impliesthat transcription normally proceeds throughout loop length, though this statementneeds to be qualified to take account of both electron- and light-microscope evidencefor intercalary termination and initiation zones within some loops (Angelier & Lacroix,1975; Scheer, Franke, Trendelenburg & Spring, 1976).

There are a few loops which show an exceptional labelling pattern. Thus thegiant granular loops near the left end of chromosome XII of Tritttrus cristatus cristatusare only labelled in the dense region, nearby the thin insertion, if preparations aremade immediately following a short-term provision of pHJuridine. As the period ofavailability of radioactive precursor is increased, so the proportionate labelled lengthof the giant granular loops increases (Gall & Callan, 1962). These loops are thereforeexceptional in that the zone where transcription occurs is only a part of the extendedloop axis; the transcribed RNA and its associated protein continues thereafter to

• Present address: MRC Clinical and Population Cytogenetics Unit, Western GeneralHospital, Crewe Road, Edinburgh EH42 XU, Scotland.

28o 5. E. Hartley and H. G. Callan

traverse a 'silent' region of loop axis before being shed into the nuclear sap at thethick loop insertion. The giant granular loops have come to be known as 'sequentiallylabelling', and a few loops labelling in similar fashion have been found in otherurodeles: a pair near the left end of chromosome XI of Notophthalmus (Triturus)viridescens (Gall, 1963) and one of the pairs of giant fusing loops on chromosome Xof Triturus marmoratus (Nardi, Ragghianti & Mancino, 1972).

In experiments where a labelled RNA precursor is supplied to oocytes for a longtime period - several days - the status of labelling of a given lampbrush loop reflectsnot only the rate of transcription, and the specific activity of the precursor in itspool, but also what has happened to the transcribed RNA thereafter, e.g. processingof transcript (Old, Callan & Gross, 1977). In short-term incorporations, on thecontrary, the status of labelling directly reflects the rate of transcription. In Notoph-thalmus viridescens, Gall (unpublished observations) noticed that the rate of incor-poration of [3H]uridine into the RNA of the giant loops on chromosome II wasmarkedly lower than incorporation by the generality of smaller, ordinary loops,though the giant loops are normal in the sense that they label uniformly throughouttheir lengths in short-term incorporation experiments. This might mean that therate of RNA transcription on the giant loops is lower than elsewhere, but it mightalso be occasioned by a lower uracil content of the giant loops' RNA as comparedwith those RNAs transcribed on other loops. We decided to attempt to distinguishbetween these two possibilities by studying the relative rates of incorporation of all4 RNA precursors when administered separately.

We had a further motive for attempting this study. The giant loops of chromosomeII of N. viridescens are exceptional in resisting digestion by the restriction endo-nuclease Hoe, whereas all the other loops of this organism's chromosome complement,and the main chromosome axes, are sensitive (Gould, Callan & Thomas, 1976). Wethought that this refractory behaviour might be reflected in an unusual base ratioof the giant loops' RNA.

It would be difficult to collect sufficient unlabelled RNA from the giant loops,uncontaminated by RNA from other loops, for a direct determination of base ratio.It would be well nigh impossible to compute base ratio from labelled giant loops'RNA for lack of knowledge of the specific activities of the 4 precursors when theyhave entered oocytes. But the problem presented by unknown specific activitiescan be circumvented if the labelling levels of different loops from single oocytes arecompared with one another, for these various loops will have been transcribingRNA from a common precursor pool. In autoradiographs the giant loops of chromo-some II can be readily recognized, but not so other loops. Our comparisons havetherefore been made between the labelling levels of the giant loops, and pooledinformation from several other unidentified loops, on the assumption that thegenerality of these other loops carry RNAs whose base ratios, taken overall, wouldconform to the base ratio of total lampbrush chromosomal RNA of N. viridescens.This latter information is available from the studies of Edstrom & Gall (1963).

Transcription on newt lampbrush chromosomes 281

MATERIALS AND METHODS

Adult females of the American red spotted newt, Notophthalmus (Triturus) viridescens wereobtained from Lee's Newt Farm, Oak Ridge, Tennessee, and kept in the laboratory at 18 °C.To obtain oocytes a newt was anaesthetized with MS222 (Sandoz), a small incision made inthe abdominal wall, and an ovary removed. Thereafter the newt's body wall was sutured, sothat the animal was ready for a similar operation the following day.

Administrative of radioactive precursors

Oocytes were incubated in vitro with the radioactive precursors of RNA taken separately.[3H]adenine (No. 408130B, 23 Ci/mM), ["HJcytidine (No. 401571, 27-3 Ci/mM), [3H]guanosine(No. 380302, 19 Ci/mM) and ['HJuridine (No. 345471, 25 Ci/mM) were obtained from theRadiochemical Centre, Amersham. For the experiments 50 /tCi of a single precursor was putinto an individual embryo cup and allowed to dry down overnight. Half of an ovary, with alittle coelomic fluid, was placed in this embryo cup and twirled about so as to put the precursorinto solution. A Vaselined glass lid was placed over the embryo cup to reduce evaporation.Periodically the lid was lifted, and the half ovary twirled about in an endeavour to distributethe precursor uniformly amongst the oocytes. Incubation proceeded for 4 h at room tem-perature, about 18 °C. At the end of incubation the embryo cup was placed on ice. The otherhalf ovary was put with a different precursor and similarly treated. The following day thesecond ovary was removed and divided into 2 halves, and these likewise placed with theremaining 2 precursors. In this manner oocytes, all from one animal, could be compareddirectly in their capacities to incorporate the individual precursors.

Centrifuged lampbrush chromosomes

The techniques used for isolating oocyte nuclei and removing nuclear membranes wereas described by Callan & Lloyd (1960), and centrifuged preparations of lampbrush chromo-somes were made as described by Old et al. (1977), except that the chromosome-dispersingmedium contained 05 x io"4 M CaCl2 and the centrifugation was done at 3000 rev/min(1055 g). Oocytes of diameters 0-6-0-8 mm were chosen for chromosome isolation, a sizeat which the lateral loops are well extended in iV. viridescens, yet capable of producing pre-parations with the bivalents reasonably spread apart from one another.

Centrifuged preparations of lampbrush chromosomes were fixed in a 4 % solution of com-mercial formalin for 15 min or longer. Each coverslip, with its attached chromosomes, wasseparated from its ring cell, washed in running filtered tapwater to remove the formalin,placed for 5 min in freshly prepared ice-cold 5 % trichloroacetic acid for extraction of anyunincorporated precursors, again washed in running filtered tapwatei, followed by distilledwater. The covetslip preparations were then taken through an ethanol series to 2 changesof xylene to remove the wax, 2 changes of acetone to remove the xylene, and air-dried. Thecoverslips were now attached to 3 x 1 in. (7-6 x 2-5 cm) microscope slides with a minimumamount of Canada balsam, the chromosomes being uppermost and exposed.

As controls, several slides were digested with ribonuclease (RNase). These slides wereplaced in a solution of RNase-A from bovine pancreas, obtained from the Sigma ChemicalCompany, in o-oi M phosphate buffer at pH 6-o, at a concentration of 01 mg/ml. Theywere digested for 1 h at 40 °C, then washed in distilled water and air-dried.

Autoradiography

The preparations were filmed with Kodak NTB-2 dipping emulsion diluted with distilledwater to half its original strength. After filming, the slides were stored in a refrigerator duringexposure. The autoradiographs were developed for 2-5 min in freshly prepared Kodak D-19at 20 °C, washed in distilled water, fixed in Kodak Unifix, washed in filtered tapwater followedby distilled water, and air-dried. The autoradiographs were studied by placing a drop ofdistilled water and then a coverslip over the preparations and then observing them underphase-contrast optics ( x 40) using a Zeiss Standard WL microscope.

282 S. E. Hartley and H. G. Callan

For photography the preparations were stained in 0-4 % toluidine blue for 45 min, rinsedwith distilled water, and air-dried. The preparations were photographed through an Ilford 404green filter and a Zeiss broad-band green interference filter using a x 40 apochromatic oil-immersion objective and bright-field optics. Immersion oil placed directly on the preparationwas removed by petroleum ether followed by acetone and air-drying.

Analysis of autoradiographs

Oocytes from 5 animals were allowed to incorporate the 4 individual precursors. Test slideswere developed at intervals to ascertain appropriate exposure times, which varied from 7 to17 days in the cases of f'HJadenine, [sH]cytidine and pHJuridine; to achieve a comparablelevel of labelling for ftTJguanosine it was necessary to expose the slides for between iz and26 weeks.

The developed autoradiographs were analysed in the following way. Using a Reichertcamera lucida attached to the Zeiss microscope, drawings were made of the giant loops onchromosome II and a number of unidentified ordinary loops which happened to lie con-veniently such that the total length of ordinary loops drawn (average 850 fim) exceeded thatof the giant loops (average 550 fim) in each preparation analysed. The number of silver grainswas counted over the drawn loops. From a series of direct observations the average widthof the giant loops in centrifuged preparations was taken to be 165 /tm and that of the ordinaryloops, 083 /tm. A correction for background was made by multiplying the assumed looparea by a background correction factor to calculate the number of grains over the loops whichcould be attributed to background. This figure was then subtracted from the original graincount to give the corrected grain count. The background correction factor was derived bycounting the number of grains in 6 different areas on each slide, selected at random; eacharea measured 900 /tm1. Grains were counted within the area and lying on one vertical andone horizontal line bordering each area, the same lines being used for each area counted.The number of grains pel fim1 was calculated for each preparation and this was the back-ground correction factor. The corrected grain count was then divided by th measured looplength to derive the number of grains per /tm. A ratio of grains per /tm. for giant loops tograins per /tm for ordinary loops was then calculated.

OBSERVATIONS

Autoradiographs following [3H]adenine, [3H]cytidine, [3H]guanosine and PH]-uridine incorporation by the giant loops on chromosome II are shown in Figs. 1-4.RNased preparations were uniformly negative. When compared with the ordinaryloops, the level of labelling on the giant loops is in all cases significantly lower(Table 1). In the cases of pHJadenine (Fig. 1; column 3 of Table 1) and [3H]uridine(Fig. 4; column 9 of Table 1) the ordinary loops are twice as heavily labelled as thegiant loops; with [3H]cytidine (Fig. 2; column 5 of Table 1) the ordinary loops are

Figs. 1-4. Photographs of autoradiographs stained with toluidine blue and selectedso as to include the giant loops on chromosome II.

Fig. 1. PHJadenine incorporation. The giant loops are weakly labelled. Thewell labelled long loops with relatively little RNP matrix at the bottom of thephotograph are not the giant loops; they are representative of the 'generality' ofordinary loops with whose level of labelling that of the giant loops has been compared(17 days exposure).

Fig. 2. fFfJcytidine incorporation (17 days exposure).Fig. 3. PFTIguanosine incorporation (26 weeks exposure).Fig. 4. [3H]uridine incorporation (17 days exposure).

Transcription on newt lampbrush chromosomes

* • * 3<

284 S. E. Hartley and H. G. Callan

Table 1. Levels of labelling of the giant loops on chromosome II ofN. viridescens relative to those of ordinary loops

Newt

1

2

3

4

5

Meanrelativelevels

Oocyte

1

2

34

1

2

345

1

2

34

1

2

34

1

2

345678

[3H]adenine

o-490-480 6 4

o-47

0-40

o-470 6 70-670 8 7

0-850-650 4 4

o-43

° 4 30-42

o-340-37

0-300-520 6 8

o-590 2 50-51

°S70-32

0 5 1

±o-io

Oocyte

1

2

345678

1

2

3456

1

2

34561

2

34

1

2

34567

["PTJcytidine

0-871 080 9 70 9 6

076O790 8 2086

0-82

0 8 60 9 60 9 80 9 0

078

0 8 10 8 80 8 30 8 1

o-880-82

0 9 9

o-880 9 80 7 0

068I-OI

0 8 90-91

o-97o-970 9 2

088±008

Oocyte

1

2

345

1

2

1

2

34S

1

2

34S1

2

3456789

[3H]guano-sine

0-330-27

0 1 6

o-S70-29

0 1 8

0 2 8

0-14

0-180 1 80-26

0 1 8

0 2 0

0-180-27

0-140 3 0

0 2 80 3 2

0-230-24

O-22

0-14

0 5 1

0 1 3

O-I9

O 2 5

±0 06

Oocyte

1

2

34567

1

2

345

1

2

34561

2

3

1

2

3456789

["HJuridine

0-48o-430-52

0 5 1

o-590-58o-53

0-480 6 30-490 7 40 5 2

0-42

cs80 6 4

0 5 1

o-6io-390 6 3

0 5 6

cs8

o-330 5 20-30

0 4 9

0-260 3 2

°-37o-340-42

0 5 1

±007

only slightly more heavily labelled, but with [3H]guanosine (Fig. 3; column 7 ofTable 1) the ordinary loops are 4 times as heavily labelled as the giant loops.

Despite the precursors having reasonably comparable initial specific activities, forcomparable silver grain densities the preparations in which [3H]guanosine was usedas the RNA precursor required about 10 times longer exposure than the preparationsinvolving the other three precursors. The lengthy exposures required for the [3H]-guanosine preparations might be attributed to a much higher guanosine pool in the

Transcription on newt lampbrush chromosomes 285

oocytes, or slower entry of [3H]guanosine into the oocytes. In an attempt to reduceexposure times for this RNA component the experiments were repeated on oneanimal using [3H]guanine sulphate (No. S16325, 8-4 Ci/mM) instead of pHJguanosine,but after 6 weeks exposure it was evident that these preparations were essentiallyunlabelled. In any event the longer exposures for the [3H]guanosine preparationsis not of particular consequence, for as already mentioned the precursor pool problemhas been essentially circumvented by making comparisons within single preparationswhere all chromosomes must have been transcribing RNA from a common pool.

Table 2. Base ratio of the RNA transcribed from the giant loops (GLs)on chromosome II of N. viridescens

Base

AdenineGuanineCytosineUracil

Rate of 1

Levels of labellingof GLs relativeto the generality

of ordinary loops,%

5124-58851

Base ratio of thegenerality of lamp-brush chromosomal

RNA ofN. viridescens

(from Edstrom &GPII, 1063), %

2 6 620-52 3 8

2 9 3

RNA transcription from the GLs ascompared with that from theordinary loops

generality of

Base ratio of GLs'RNA weighted inaccord with the

overall compositionof chromosomal

RNA

i 3 S5-o

2 0 91 4 9

543

Base ratioof GLs' RNA,

0'0

2 4 99 2

38-527-4

Edstrom & Gall (1963) isolated total chromosomal RNA from A . viridescens andanalysed its base composition (column 3 of Table 2). As we have determined theratio for the level of label in the giant loops with respect to various unidentifiedordinary loops, the base ratio of the RNA transcribed by the giant loops can becalculated. This is done by multiplying the ratio for the level of label incorporatedfor each precursor by the relative amount of that residue as determined by Edstrom &Gall for the total chromosomal RNA (column 4 of Table 2). The 5th column ofTable 2 shows the base ratio of the giant loops' RNA expressed as percentages.There is an extreme imbalance between guanine (9-2%) and cytidine (38-5%),which in turn implies a comparable imbalance between the transcribed and non-transcribed strands of the DNA in the giant loops' axes.

It will be recalled that Gall originally noted the relatively low rate of incorporationof [3H]uridine by the N. viridescens giant loops, but this observation by itself didnot allow the inference that the rate of RNA transcription on these loops is lowerthan elsewhere; the giant loops' RNA might merely be relatively deficient in uridine.The results presented here confirm Gall's observation but go further and showthat the overall rate of RNA transcription on the giant loops is genuinely lower thanelsewhere, little more than half the average (Table 2).

19 CEL34

286 S. E. Hartley and H. G. Callan

DISCUSSION

These studies have revealed two unusual features of the giant lampbrush loops onchromosome II of N. viridescens. First, they synthesize RNA at a relatively low rate.This might be occasioned by a low rate of RNA polymerase attachment to thezone(s) where transcription begins, and/or a slow rate of polymerase movement inthe course of transcription. Though these two factors are doubtless interrelated, wewould attach more significance to the latter.

There are various reasons for supposing that the extended lengths of lampbrushlateral loops depend inter alia, on the density of RNA polymerase molecules attachedto loop axis (see Franke et al. 1976). Thus, for example, when RNA polymerasesare caused to detach, with their accompanying transcript RNP, by poisoning newtoocytes with actinomycin D, the contraction of lateral loops is dramatic (Snow &Callan, 1969) and in the course of recovery thereafter, coincident with the reacquisitionof capacity to synthesize RNA, comes re-emergence of lateral loops. Pelling (1972)likewise maintains that it is the density of RNA polymerase molecules on the puffsand Balbiani rings of polytene chromosomes which primarily determines the sizeof these transcriptionally active regions.

Now Hartley (1977) has found that when N. viridescens females are kept at 4 °Cfor a few days, their lampbrush loops become shorter, but this shortening is pro-portionately less in the case of the giant loops on chromosome II than for the generalityof loops. Cold treatment reduces the overall metabolic activity of newts, and mightbe expected to reduce both the rate of RNA polymerase movement on the loops, andalso the rate of polymerase attachment. Initially, cold treatment would not beexpected to affect polymerase density on the loops, and hence loop length; but inthe course of time, as transcripts which were already in course of synthesis detach,the effect of reduced rate of polymerase attachment should be expressed as a loweringof polymerase density on the loops, and hence loop shortening.

There is no a priori reason to anticipate a disproportionate reduction in the rateof attachment of polymerases to the giant loops, or rate of movement of transcriptson these loops, as a result of lowering environmental temperature. However, if therate of transcript movement on the giant loops is intrinsically low, and given thatthe transcriptional units on the giant loops are exceptionally long, these are twoconcordant factors which would be expected to combine to produce disproportionatelyless shortening of the giant loops, as compared to the generality, after a few daysexposure to low temperature.

There is another reason, and an obvious one, for anticipating a slower rate oftranscript movement on the giant loops than elsewhere. As just mentioned, thesegiant loops comprise exceptionally long transcriptional units, and the very bulk oftheir transcript RNP must be more of an impediment to polymerase movementthan RNP transcripts on other, shorter loops.

The other unusual feature to have emerged from this study is the gross baseimbalance in the giant loops' RNA transcript. As far as we are aware the only otherbase ratio determinations for RNAs as transcribed from particular chromosome

Transcription on newt lampbrush chromosomes 287

regions are those of Edstrom & Beermann (1962) for the Balbiani ring RNAs on the4th (smallest) chromosome of Chironomus tentans. They found, inter alia, that theRNA transcribed on Balbiani ring 2 has a base ratio of approximately: adenine 38,guanine 20-5, cytosine 24-5, and uracil 17, the imbalance between adenine and uracilimplying that there are more than twice as many thymine residues on the transcribedas on the non-transcribed strands of the Balbiani ring's DNA. In the case of thegiant loops' RNA which we have presented the base imbalance is even greater, buthere it primarily concerns guanine and cytosine, not adenine and uracil. Withguanine at 9-2 and cytosine at 38-5 in RNA, there must be about 4 times as manyguanine residues on the transcribed as on the non-transcribed strands of the giantloops' DNA.

Before a detailed investigation of grain counts over the giant loops of N. viridescenshad been made, it appeared as though they might be transcribing an RNA entirelylacking in guanine. If this had been substantiated it would at once have explainedthe results obtained by Gould et al. (1976) following their observations on the effectsof restriction endonucleases on lampbrush chromosomes. Gould et al. found thatwhile the giant loops on chromosome II of N. viridescens are cut by several nucleasesthe giant loops, and these alone, are resistant to digestion by Hae from Haemophilus

aegyptius, which recognizes the specific sequence Although the relative

abundance of guanine residues on the transcribed as opposed to the non-transcribedstrand of the DNA axis significantly diminishes the chance of encountering the

CC GCsequence ->->, in a random base order (from once in every 256 base pairs if all

4 bases are equally abundant to roughly once in every 800 base pairs) this is clearlyinsufficient to account for the indigestibility of the 200- to 3oo-/«n-long giant loops.

Evidently the giant loops on chromosome II of N. viridescens, with their low rateof RNA synthesis, the grossly unbalanced base ratio of their RNA, and their lackof sites sensitive to endonuclease Hae, are unusual, and the function of the RNAwhich they transcribe remains a problem for the future.

This work was supported by grant B/SR/88418 from the Science Research Council.

REFERENCES

ANGELIER, N. & LACROIX, J.-C. (1975). Complexes de transcription d'origines nucl£olaire etchromosomique d'ovocytes de Pleurodeles waltlii et P. poireti (Amphibiens, Urodeles).Chromosoma 51, 323-335.

CALLAN, H. G. & LLOYD, L. (i960). Lampbrush chromosomes of crested newts Trituruscristatus (Laurenti). Phil. Trans. R. Soc. Ser. B 243, 135-219.

EDSTROM, J. E. & BEERMANN, W. (1962). The base composition of nucleic acids in chromosomes,puffs, nucleoli and cytoplasm of Chironomus salivary gland cells. J. Cell Biol. 14, 371-380.

EDSTROM, J. E. & GALL, ] . G. (1963). The base composition of ribonucleic acid in lampbrushchromosomes, nucleoli, nuclear sap and cytoplasm of Triturus oocytes. J. Cell Biol. 19,279-284.

FRANKE, W. W., SCHEER, U., TRENDELENBURG, M. F., SPRING, H. & ZENTGRAF, H. (1976).

Absence of nucleosomes in transcriptionally active chromatin. Cytobiologie 13, 401-434.

19-2

288 S. E. Hartley and H. G. Callan

GALL, J. G. (1958). Chromosomal differentiation. In A Symposium on the Chemical Basis ofDevelopment (ed, W. D. McElroy & B. Glass), pp. 103-135. Baltimore: Johns HopkinsPress.

GALL, J. G. (1963). Chromosomes and cytodifferentiation. In Cytodifferentiation and Macro-molecular Synthesis (ed. M. Locke), pp. 119-143. New York: Academic Press.

GALL, J. G. & CALLAN, H. G. (1962). H'-uridine incorporation in lampbrush chiomosomes.Proc. natn. Acad. Sci. U.S.A. 48, 562-570.

GOULD, D. C, CALLAN, H. G. & THOMAS, C. A. Jr. (1976). The actions of restriction endo-nucleases on lampbrush chromosomes. J. Cell Sci. 21, 303-315.

HARTLEY, S. E. (1977). Studies on the Lampbrush Chromosomes of the American Newt Notoph-thalmus (Triturus) viridescens. Ph.D. Thesis, University of St Andrews, St Andrews,Scotland.

IZAWA, J., AT.I.FREY, V. G. & MIRSKY, A. E. (1963). The relationship between RNA synthesisand loop structure in lampbrush chromosomes. Proc. natn. Acad. Sci. U.S.A. 49, 544-551.

NARDI, I., RAGGHIANTI, N. & MANCINO, G. (1972). Characterization of the lampbrushchromosomes of the marbled newt Triturus marmoratus. Chromosoma 37, 1-22.

OLD, R. W., CALLAN, H. G. & GROSS, K. W. (1977). Localization of histone gene transcriptsin newt lampbrush chromosomes by in situ hybridization. J. Cell Sci. 27, 57^79.

PELLING, C. (1972). RNA synthesis in giant chromosomal puffs and the mode of purring.FEBS Symp. 24, 77-89.

SCHEER, U., FRANKE, W. W., TRENDELENBURG, M. F. & SPRING, H. (1976). Classification ofloops of lampbrush chromosomes according to the arrangement of transcriptional complexes.J. Cell Sci. 22, 503-519.

SNOW, M. H. L. & CALLAN, H. G. (1969). Evidence for a polarized movement of the lateralloops of newt lampbrush chromosomes during oogenesis. J. Cell Sci. 5, 1-25.

(Received 15 May 1978)