Upgraded Expression of 5S rRNA Preludes theProduction of Fibroin by Spider Glands

EDWIN VAZQUEZ, GERADO ARROYO, IVAN J. CAJIGAS, andGRACIELA C. CANDELASn

University of Puerto Rico, San Juan, Puerto Rico 00931

ABSTRACT We have developed the large ampullate glands of the orb-web spider Nephilaclavipes as a model system in which to study the production of a tissue-specific secretory protein.Through simple manipulations, the glands’ fibroin production can be practically abolished andsubsequently elicited into high levels of synthesis through a mechanical stimulus applied to theorganism. The tissue specific responses evoked by the stimulus can be monitored through time-course studies. The latter have revealed an orchestrated series of tissue and time specificmacromolecular syntheses, which optimize the glandular tissues with components of the proteinsynthesis machinery. This work shows the upgraded accumulation of 5S rRNA in the glands asresponse to the stimulus within the earliest of the prelude events. Further enquiries on thisaccumulation must be conducted at the level of differential gene expressions, a chore we haveinitiated. A DNA fragment containing a single copy 5S rRNA gene has been isolated, cloned,sequenced, and transcribed in a cell-free system. We enclose a discussion on the similarity betweenthe genomic organization of this gene to that of a 5S rRNA gene of Bombyx mori. Our studies haverevealed a considerable number of similarities in the silk production strategies of Nephila clavipesand the silkworm Bombyx mori, some of them rather unusual. J. Exp. Zool. 298A:128–133, 2003.r 2003 Wiley-Liss, Inc.

INTRODUCTION

In the first article appearing in this journal onthe large ampullate glands of the orb-web spiderNephila clavipes (Candelas and Cintron, ’81), weconcluded that these glands ‘‘comprise a favorablesystem in which to study the synthesis of a largesecretory protein and its control.’’ Subsequentresearch from our laboratory has confirmed ourprediction. The glands have certainly proven theirworth as a model system. Time sequence monitor-ing of the elicited process via a stimulus applied tothe organisms, have revealed that the productionof the glands’ full size tissue-specific product is theculmination of a series of tissue and time-specificdiscrete waves of macromolecular syntheses(Candelas and Lopez, ’83; Candelas et al., ’90).These waves of activities, all of which generateRNAs, optimize the gland’s tissues for thedemanding feat of a daily production of high levelsof the very large fibroin. The fibroin is known to bein constant demand during the entire life of theadult female and is critical for a number of its lifefunctions (Peakall, ’64). Sixty minutes prior to theproduction of the fibroin, which peaks after ninetyminutes of incubation of the stimulated glands, a

wave of template RNA is generated (Candelas andLopez, ’83). The latter is followed by a discretewave of selected tRNAs syntheses, which adap-tively shift the gland’s tRNA population for thedecodification of a template rich in fibroin’spreponderant amino acids (Candelas et al., ’90).

We are currently identifying small RNAs gen-erated during the earliest of the activities preced-ing the wave that provides the mRNA by fifteenminutes. This event enriches the glands with aconsiderable variety of small RNAs (Candelaset al., ’93). Thus far, a rather thick band of atissue-specific and fibroin synthesis correlatedalanine tRNA has been identified (Candelas et al.,’90). Although the role of this isoform remains tobe elucidated, it is reasonable to assume that itplays a role in the fibroin production strategy,since it is also a component within the strategy of

Grant sponsor: National Institutes of Health; Grant number:RR003641 to G.C.C. and Institutional Funds.

nCorrespondence to: Graciela C. Candelas, Department of Biology,University of Puerto Rico, P.O. Box 23360, University Station, SanJuan, Puerto Rico 00931–3360. E-mail: gcandel@rrpac.upr.clu.edu

Received 30 October 2002; Accepted 12 March 2003Published online in Wiley InterScience (www.interscience.wiley.

com) DOI:10.1002/jez.a.10268

r 2003 WILEY-LISS, INC.

JOURNAL OF EXPERIMENTAL ZOOLOGY 298A:128–133 (2003)

fibroin production of the silkworm Bombyx mori(Sprague et al., ’77).

The aim of the present work includes theidentification of another small RNA selectivelyaccumulated during this first wave of the stimu-lus-elicited events. Two bands pertaining to 5SrRNA have been identified, both with similardegrees of accumulation, as a function of thestimulation of the glands into fibroin synthesis.Our first search into the spider genome for thesegenes has yielded a DNA fragment containing asingle copy gene. The gene has been cloned,sequenced, and transcribed in a cell-free system,as reported subsequently. The similarities ingenomic organization displayed by the Nephilaclavipes 5S rRNA and that of a 5S rRNA Bombyxmori gene (Morton and Sprague, ’82) are dis-cussed. We also stress the impressive degree ofsimilarity between the sequence of the 5S rRNAgene coding region and the 5’ flanking regulatoryelements of these two silk producing, but highlydivergent, organisms.

MATERIALS AND METHODS

Experimental Organisms

Collection and handling of Nephila clavipesmature females, as well as the methods toinactivate and/or mechanically stimulate synthesisof fibroin by the large ampullate glands, havepreviously been described (Candelas and Lopez,’83). This reference also offers details on thedissection of the glands and the maintenance oftheir metabolic activity in a simple culture medium.

Isolation of total RNA

The large ampullate glands of unstimulated andmechanically stimulated spiders were excised andincubated in 1X SSC for either 15 or 45 minutes.RNA isolation was performed following a modi-fication of the rapid one-step procedure ofChomczynski and Sacchi (’87) using the RNAgentsTotal Isolation System (Promega, Madison, WI).

Northern analyses

RNA samples were fractionated in a 10%polyacrylamide bis 29:1, 7M urea, 1X TBE(89mM tris, 89mM boric acid, 2mM EDTA, pH8.0) gel at 800V for 24 hours at 41C. They wereelectrotransfered unto Zetabind nylon membranes(CUNO, Meriden, CT) under conditions previouslydescribed (Candelas et al., ’90). Membranes wereprehybridized in a solution containing 5X SSC,

0.3% SDS and 5X Denhardt’s reagent. For condi-tions of high stringency, prehybridizations, hybri-dizations, and washes were performed at 601Cinstead of at 501C. Hybridization was conductedovernight by adding 10ml of prehybridizationsolution and 1� 106 cpm/ml of a probe. Theoligonucleotide CTG ACT ACG CCC GAC GTTGCT TGA CTT CGG TGA TCG GAC GAG AACCGG TGT TTT CAA CGT GGT ATG GAC GTTGGC complementary to the first 75nt of theBombyx mori 5S rRNA gene was used as a probe.The oligonucleotide was 32P end-labeled using theT4 Polynucleotide Kinase reaction. Gel imagesand autoradiograms were scanned using a GS-800calibrated densitometer (Bio-Rad, Hercules, CA).Scanned image files were inspected for artifactsand submitted to volume densitometric analysesusing the Quantity One v.4.3 quantitation soft-ware (BioRad, Hercules, CA). Normalization of thehybridization signals across samples was deter-mined using the 18S rRNA levels as an endogen-ous internal control to measure backgroundhybridization levels. Measures of 5S rRNA levelswere expressed as a ratio of the measures of 5SrRNA: 18S rRNA volumes.

Southern blots

Genomic DNA, isolated through standard meth-ods, was digested under the established conditionswith a variety of restriction endonucleases (NewEngland BioLabs, Beverly, MA). Samples wereseparated through 1% agarose gel electrophoresisin a Tris-borate buffer system and transferred to aZetabind nylon membrane (CUNO, Meriden, CT).Membranes were prehybridized and hybridized asdescribed for Northern blots. High stringencywashes were performed using 1X or 0.1X SSC.The probe used for the recognition of 5S rRNAgenes was the same used in Northern blots.

Cloning

Based on the Southern analyses, 50mg of an AvaI digest sample of Nephila clavipes genome wasfractionated via agarose gel electrophoresis. Thearea of the gel where fragments generatedhybridization signals was excised. The DNA waseluted and purified using Gene Clean (Bio 101,Vista, CA) and partial genomic libraries wereprepared in plasmid pGem 7Zf (�) (Promega,Madison, WI). Libraries were screened by in situcolony hybridization under the conditions pre-viously described for Northern blots. Clonesemitting a positive signal for 5S rRNA gene were

ELICITED SYNTHESIS OF SPIDER FIBROIN 129

submitted to automated laser fluorescent sequen-cing using the Cy5 Auto-Read sequencing kit(Amersham-Pharmacia Biotech, New York, NY).Sequence analysis was performed using NIH-NCBI analysis tools. The sequence of clonepNC18F containing a Nephila clavipes 5S rRNAgene and a repetitive element gene (Nc1) wassubmitted to Gene Bank. Their accession numbersare U52004 and U54613 respectively.

Cell free transcription

Clone pNC18F was submitted to transcriptionin a heterologous crude nuclear extract derived inour laboratory from the posterior silkglands ofBombyx mori. Properly optimized for Nephilaclavipes genes, the extracts have been shown torender faithful transcripts (Cintron et al, 1999).Transcription reactions were conducted under theconditions described in the cited reference. Theproducts were fractionated in a 12% polyacryla-mide (19:1 acrylamide/bis, 7M Urea, 0.1%SDS/1XKBT; KBT is 49.5mM tris, 49.5mM boric acid and0.78mM EDTA) gel and run at a constant voltageof 150V for 20 hours at room temperature. The

transcripts were detected by autoradiography. Thenucleic acid marker (Invitrogen Life Technologies,CA) was radiolabeled using T4 polynucleotidekinase. Recognition of the 5S rRNA gene tran-scripts was obtained through Northern blotanalyses of unlabeled transcription productsagainst the 32P labeled probe described previously.

RESULTS

The first set of experiments of this work wasdesigned to ask if enrichment with 5S rRNAprecedes the stimulated production of fibroin inthe large ampullate glands of the orb web spiderNephila clavipes during the first wave of stimuluselicited activities. The data shows the upgradedaccumulation of 5S rRNA as an activity precedingthe evoked fibroin production. Panel A in Figure 1shows the mobility pattern of the RNA speciesunder denatured conditions. These were subse-quently submitted to Northern hybridization, asdescribed in the methods. Two bands wererecognized as 5S rRNA (Fig. 1B). Lane 2 (PanelsA and B) corresponding to RNA isolated fromstimulated glands incubated during 15 minutes

Fig. 1. Fractionation of total RNA from the sourcesindicated in each panel and corresponding Northern Blotsagainst a 5S rRNA probe, as described in methods. Panel A.RNA from glands of non-stimulated spiders, lane 1. RNAsfrom the glands of stimulated organisms at 15 and 45 minutesafter stimulation in lanes 2 and 3 respectively. 18S rRNA,indicated by arrow, has been used as an endogenous internalcontrol. Panel B. The corresponding Northern hybridization,

showing two isoforms recognized by the probe under highstringency conditions, and both displaying upgraded accumu-lation of 5S rRNA in response to the organism’s stimulationinto fibroin synthesis. Panel C. RNAs from non-glandulartissues at 15 minutes after stimulation and from non-stimulated organisms, lane 1 and 2 respectively. Panel D.Northen hybridization, reveals two isoforms in non-glandulartissues.

E. VAZQUEZ ET AL.130

show an increase in the accumulation of the 5SrRNA species when compared with non-stimu-lated glands (Lane 1, Panels A and B). The 18SrRNA (labeled in Fig. 1A) was used as an internalcontrol since this RNA has been found to beinvariable in most tissues thus far examined(Spanakis, ’93). The 5S rRNA shows no increasein the non-glandular tissue, which was subjectedto the same experimental conditions (Panel C).Two signals for 5S rRNA are seen in non-glandular tissue (Panel D), implying that neitherof the two forms seen in the stimulated glands aretissue-specific, contrary to the case in the twoalanine tRNAs (Candelas et al., ’90).

Densitometric analyses conducted to estimatethe magnitude of the accumulation of the 5S rRNAspecies in response to the gland’s stimulation areseen in Figure 2. The uppermost band hasincreased 31%, the lowermost displays an increaseof 44% at 15 minutes after glands stimulation.

The previous data indicate that the expressionof 5S rRNA gene(s) has been enhanced by thefibroin producing stimulus applied to the organ-ism. Further enquiries on the latter must bescrutinized at the level of gene expression andtheir cis-regulatory elements. Thus far we haveisolated one 5S rRNA gene. The gene representsa single copy, inasmuch as it was isolated from a2 kb spider DNA fragment and no neighboring5S rRNA genes were found within the fragment.

The gene has been cloned (pNC18F) and tran-scribed in a cell-free system, as described. Arepetitive DNA element (Nc1) is also present inthis clone.

Figure 3 (lane 2) shows the products of the cell-free transcription of the cloned spider 5S rRNA.Transcription of pNC18F yielded a product ofestimated molecular size of 120 nt. The identifica-tion of the transcription product as a 5S rRNA isshown in Figure 4. A 5S rRNA transcript was

Fig. 2. Densitometric analysis of the 5S rRNA accumula-tion at glandular tissues. Hybridization signals obtained inPanel B (lanes 1, 2 and 3) were subjected to densitometricanalysis as decribed in the methods. Values of 5S rRNA levelswere expressed as a ratio of the measures of 5S rRNA: 18SrRNA volumes.

Fig. 3. Cell-free transcription of a 5S rRNA gene. Lane mdisplays a 10 nt ladder of marker RNA. Lane 1, clone pNTA1containing a Nephila clavipes tRNAAla gene used as template,Lane 2, clone pNC18F, containing a spider 5S rRNA gene anda repetitive element sequence. The transcription products arepointed out by the arrows.

Fig. 4. Identification of the transcription product of clonepNC18F through Northern blot. Lane 1, clone pNC18F, usedas template. Lane 2, plasmid without insert.

ELICITED SYNTHESIS OF SPIDER FIBROIN 131

identified through Northern blot analyses ofunlabeled transcription products against the 32Plabeled 5S rRNA probe described in the methods.

DISCUSSION

Having detected differential discrete accumula-tions of small RNA molecules within two waves ofactivity preluding the elicited synthesis of fibroin(Candelas et al., ’90) we focused on the 5S rRNA.This RNA forms part of the large ribosomalsubunit of the protein machinery whose assemblyis an early event in the process. The results showtwo signals, generated by hybrids of 5S rRNAsfrom tissues and the oligonucleotide used as probe.

Since it is known that 5S rRNAs undergoprocessing (Zimmerman et al., ’93), a questionmay arise regarding the two 5S rRNA bandsrecognized via Northern hybridizations (Fig. 1,Panel 2). Are these signals generated by productsof different genes coding for isoforms, or by onegene and its processing product (Prieser andLevinger, ’91)? However, the density of the low-ermost band should be denser after 45 minutes ofincubation if this band corresponds to a product ofthe RNA processing. Thus the results from ourexperiments argue against the latter.

Consonant with our results and supporting thatthe two hybrids contain products of different genes(coding for isoforms), are the data on 5S rRNA genesof Bombyx mori (Morton and Sprague, ’82). Theirstudies show that there are at least three 5S rRNAgenes coding for different isoforms in their biologicalmodel and that they display differential transcrip-tional activity when tested in cell-free systems. Thiscan be the case of the 5S rRNA of Nephila clavipessince each form shows different levels of accumula-tion in both the glandular and non-glandular tissues(Fig. 1, Panels 2 and 4). However, the intensities ofthe increased accumulations of both isoforms in theglandular tissues are comparable. It is noteworthythat approximately 66% and 73% of the 5S rRNAaccumulated by the stimulated glands at 45 minutes,for the upper and lower isoforms respectively, isaccrued during the first wave of small RNA synthesis(Fig. 2). Although both isoforms were detected in thenon-glandular tissue, indicating that neither ofthe forms is tissue-specific, they were not upgradedin response to the stimulus applied, qualifyingthe obtained response as a tissue-specific activity.

5S rRNA genes in vertebrates and invertebratesdisplay both clustered and dispersed in single copyorganizations (Srivistava and Schelssing, ’91;Nederby-Nielsen et al., ’93). Thus far, one 5S

rRNA gene has been isolated in a single copyorganization as part of the 2 kb spider DNAfragment. Sequence alignment of the gene againstGenBank’s database reveals an impressive se-quence similarity (99.7%) between the spider geneand that of a Bombyx mori 5S rRNA gene(Accession # K03316). A single substitution with-in the 5S rRNA encoding sequence is detectedwithin the 120 nt molecule. This base substitutiondoes not affect the consensus sequence of theinternal control elements of this RNA polymeraseIII transcribed gene. A remarkable sequencesimilarity is also observed between the codingregion of this gene and that of two other Bombyxmori 5S rRNA genes, 96.7% to one designatedBMRN551 and 95% to BMRN55.

A high degree of similarity is also observed inthe 5’ flanking sequences of the spider andsilkworm genes. The 5S rRNA gene of Nephilaclavipes contains two identical AT rich boxeswithin the 5’ flank in the same position wherethese are located in the silkworm’s 5S rRNA gene(Morton and Sprague, ’82). These elements areknown to be essential to the transcription ofBombyx mori 5S rRNA genes.

The 5S rRNA gene generated faithful tran-scripts in a heterologous cell-free transcriptionextract. Its transcription yielded a product of anestimated molecular size identical to that of the 5SrRNA gene product of a variety of organisms thusfar reported (Morton and Sprague, ’82; Szymanskiet al., ’97). It is reasonable to assume that theefficiency of transcription (Fig. 3) reflects the highdegree of similarity within the upstream flankingsequences of the 5S rRNA gene of Nephila clavipesto that of Bombyx mori. The band of less intensityobserved at þ/� 150 nt could pertain to thetranscript of the Nc1 gene in clone pNC18F, sincea transcript of 160 nt for a Bombyx mori repetitivesequence has been reported (Wilson et al., ’88).

The high degree of similarities between thecoding sequences of 5S rRNA genes of Nephilaclavipes and Bombyx mori is interesting from anevolutionary perspective. The structure of 5SrRNA has been used traditionally as a molecularmarker of relationships between organisms due toits resistance to evolutionary change (Xiang-Ronget al., ’82; Szymanski et al., 2002). It is remarkablethat the 5S rRNA of the spider Nephila clavipesdiffers from the silkworm Bombyx mori moleculein only one position whereas the 5S rRNA ofanother silkworm, Philosemia cynthia ricini,differs in nine positions from that of Bombyx mori(Xiang-Rong et al., ’82).

E. VAZQUEZ ET AL.132

Of evolutionary interest is also the fact that thespider 5S rRNA gene has been found to be in closeproximity to another pol III gene, a repetitiveDNA element. The organization is similar to thatof a 5S rRNA gene from Bombyx mori. Thisrepetitive element, designated Nc1, exhibits an88.9% similarity to Bm1, a Bombyx mori repetitiveelement (Wilson et al., ’88).

Studies on species-specific control signals couldreveal mechanisms that may have been preservedamong different taxa and should prove to be ofinterest in establishing evolutionary relationshipsbetween uniramians and arachnids, organismsthat are assumed to have diverged more than400 million years ago. Of interest are the facts thatin synthesizing fibroin both systems resort todiscontinuous translation (Lizardi et al., ’79;Candelas et al., ’83), and both generate a tissue-specific alanine tRNA (Sprague et al., ’77; Cha-vancy et al., ’79; Candelas et al., ’90). Comparativestudies on these two fibroin producing systemsshould be an interesting venture pursued andscrutinized at all biological levels of organization.

The large ampullate glands serve as a modelsystem for differentiation in which to search forthe network of genes involved in the production ofthe fibroin and their cis-regulatory sequences, thelong-range aim of our research. A recent windfallfrom our work is that the information accruedthrough our studies has been useful to thoseattempting to generate the silk via biotechnologiesfor both industrial and medical applications(Vollrath and Knight, 2001; Lazaris et al., 2002).

ACKNOWLEDGMENT

We would like to express appreciation toVannesa Robles for her technical assistance andfor her input and critical reading of the manuscript.

LITERATURE CITED

Candelas GC, Cintron J. 1981. A spider fibroin and itssynthesis. J Exp Zool 216:1–6.

Candelas GC, Lopez F. 1983. Synthesis of fibroin in thecultured glands of Nephila clavipes. Comp Biochem Physiol74B:637–641.

Candelas GC, Candelas T, Ortiz A, Rodrıguez O. 1983.Translational pauses during a spider fibroin synthesis.Biochem Biophys Res Com 116:1033–1038.

Candelas GC, Arroyo G, Carrasco C, Dompenciel R. 1990.Spider silkglands contain a tissue-specific alanine tRNA thataccumulates in vitro in response to the stimulus for silkprotein synthesis. Dev Biol 140:215–220.

Candelas GC, Arroyo G, Carrasco C, Carasquillo E, Plazaola A,Irizarry M. 1993. Concerted gene expressions in elicited

fibroin synthesis. In: Translational Regulation of GeneExpression 2. Ilan J, editor. New York: Plenum Press.

Chavancy G, Chevallier A, Fournier A, Garel JP. 1979.Adaptation of iso-tRNA concentration to mRNA codonadaptation in the eukaryote cell. Biochemistry 61:71–78.

Chomczynski P, Sacchi N. 1987. Single step method of RNAisolation by acid guanidinium thyocyanate-phenol-chloro-form extraction. Anal Biochem 162:156–159.

Cintron I, Capo L, Plazaola A, Arroyo G, Candelas GC. 1999.A spider tRNAAla requires a far upstream sequence elementfor expression. Gene 231:195–201.

Lazaris A, Arcidiacono S, Huang Y, Zhou J, Duguay F,Chretien N, Welsh E, Soares JW, Karatzas CN.2002. Spider silk fibers spun from soluble recombinant silkproduced in mammalian cells. Science 295:472–476.

Lizardi PM, Mandavi V, Shields D, Candelas GC. 1979.Discontinous translation of silk fibroin in reticulocyte cell-free system and intact silk glands. Proc Natl Acad Sci USA76:6211–6215.

Morton DG, Sprague KU. 1982. Silkworn 5S rRNA andalanine tRNA genes share highly conserved flanking andcoding sequences. Mol Cell Biol 2:1524–1531.

Nederby-Nielsen J, Hallenberg C, Frederiksen S, Dissing-Sorensen P, Lomholt B. 1993. Transcription of human 5SrRNA genes is influenced by an upstream DNA sequence.Nucleic Acid Research 21:3631–3636.

Peakall D. 1964. Composition, function and glandular originof silk fibroins of the spider Araneus diadematus Cl J ExpZool 156:345–352.

Preiser PR, Levinger LF. 1991. In vitro processing ofDrosophila melanogaster 5S ribosomal RNA 3’ end effectsand requirement for internal domains of mature 5S rRNA.J Biol Chem 266:7509–7516.

Spanakis E. 1993. Problems related to the interpretation ofautoradiographic data on gene expression using commonconstitutive transcripts as controls. Nucleic Acid Res16:3809–3819.

Sprague K, Hagenbuchle O, Zuniga MC. 1977. The nucleotidesequence of two silk gland alanine tRNA’s: Implicationsfor fibroin synthesis and for initiator tRNA structure.Cell 11: 561–570.

Srivistava AK, Schelssinger D. 1991. Structure and organiza-tion of ribosomal DNA. Biochimie 73:631–638.

Szymanski M, Barciszewski J, Specht T, Volker AE. 1997.Compilation of ribosomal 5S ribonucleic acid nucleotidesequences: eukaryotic 5S rRNAs. Biochim Biophys Acta1350:75–79.

Szymanski M, Barciszewska MZ, Volker AE, Barciszewski J.2002. 5S ribosomal DNA Database. Nucleic Acids Research30:176–178.

Vollrath F, Knight DP. 2001. Liquid crystalline spinningof spider silk. Nature 410:541–548.

Wilson ET, Condliffe DP, Sprague KU. 1988. Transcriptionalproperties of Bmx, a moderately repetitive silkworm genethat is a RNA polymerase III template. Molecular andCellular Biology 8:624–631.

Xiang-Rong G, Nicoghosian K, Cedergen RJ. 1982. 5S RNAsequence from the Philosamia silkworm: evidence ofvariable evolutionary rates in insect 5S RNA. Nucleic AcidsResearch 10:5711–5716.

Zimmerman RA, Dahlberg AE. 1993. Ribosomal RNA:Structure, evolution, processing and function in proteinsynthesis. Telfor, Caldwell. New Jersey.

ELICITED SYNTHESIS OF SPIDER FIBROIN 133