$6HPLTXDQWLWDWLYH3&5$VVD\IRU$VVHVVLQJ3HUNLQVXVPDULQXV,QIHFWLRQVLQWKH(DVWHUQ 2\VWHU&UDVVRVWUHDYLUJLQLFD $XWKRU V $GDP*0DUVK-XOLH'*DXWKLHU*HUDUGR59DVWD 6RXUFH 7KH-RXUQDORI3DUDVLWRORJ\9RO1R $XJ SS 3XEOLVKHGE\ The American Society of Parasitologists 6WDEOH85/ http://www.jstor.org/stable/3283856 . $FFHVVHG Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=asp. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact
[email protected].
The American Society of Parasitologists is collaborating with JSTOR to digitize, preserve and extend access to The Journal of Parasitology.
http://www.jstor.org
J. Parasitol, 81(4), 1995, p. 577-583 ? AmericanSocietyof Parasitologists1995
A SEMIQUANTITATIVE PCR ASSAY FORASSESSING PERKINSUSMARINUSINFECTIONS IN THEEASTERNOYSTER,CRASSOSTREAVIRGINICA Adam G. Marsh, Julie D. Gauthier,and Gerardo R. Vasta* of Maryland Centerof MarineBiotechnology,University BiotechnologyInstitute, 701 East PrattStreet,Baltimore,Maryland 21202 A 3.2-kb fragmentof PerkinsusmarinusDNA was cloned and sequenced.A noncodingdomain was identifiedand targetedfor the development of a semiquantitativepolymerasechain reaction(PCR) assay for the presenceof P. marinusin easternoyster tissues. The assay involves extractingtotal DNA from oysterhemolymphand using I g of that DNA as template in a stringentPCR amplificationwith oligonucleotideprimersthat are specific for the P. marinus 3.2-kb fragment.With this assay, we can detect 10 pg of total P. marinusDNA per 1 g of oyster hemocyte DNA with ethidium bromide (EtBr)staining of agarosegels, 100 fg total P. marinusDNA with Southernblot autoradiography, and 10 fg of total P. marinusDNA with dotblot hybridizations.We have used the sensitivity of the PCR assay to develop a method for estimatingthe level of P. marinus DNA in oysterhemolymphand have successfullyappliedthis techniqueto gill tissues.Oursemiquantitativeassayuses a dilution series to essentially titrate the point at which a P. marinus DNA targetis no longer amplifiedin a sample. We refer to this technique as "dilution endpoint" PCR. Using hemocytes obtained by withdrawinga 1-ml sample of hemolymph, this assay provides a nondestructivemethodologyfor rapidlyscreeninglargenumbersof adult oystersfor the presenceand quantification ofP. marinusinfectionlevels. This techniqueis applicableto othertissues(gills)and could potentiallybe appliedto DNA extracts of whole larvae or spat. ABSTRACT:
The severe disease caused by the endoparasitic protozoan Perkinsus marinus (Apicomplexa: class Perkinsea; Levine, 1978) is a major cause of mortality in the eastern oyster (Crassostrea virginica) along the Gulf of Mexico and Atlantic coast, U.S.A. A review on the epizootiology of the disease has been published by Andrews (1988). In its vegetative form as a trophozoite, P. marinus can proliferate throughout the tissues of the oyster by multiple fission, budding, or both. Eventually, mature trophozoites enlarge and then undergo rapid reductive divisions to form a hypnospore that enters the water column and sporulates to release large numbers ofbiflagellated zoospores (1,000-2,000 per hypnospore; Perkins, 1966). These motile zoospores presumably give rise to trophozoites once they infect an oyster, but neither the mechanism of infection is known, nor the oyster life stage that is the most susceptible to parasite entry. Although significant progress has been made in understanding this disease over the last 25 yr (Perkins, 1988), new methodologies need to be developed to address these gaps in our knowledge before coherent treatment and management strategies can be formulated. Diagnosis of P. marinus infections presently relies on either microscopical examination or the fluid thioglycolate assay (FTM) of Ray (1952, 1966). The FTM is generally used for screening large numbers of oysters, but the primary limitation of this technique is its inability to detect all life stages of P. marinus. The sensitivity of the polymerase chain reaction (PCR) for trace quantities of foreign DNAs in heterogeneous samples has made this technology an ideal choice for identifying infectious agents and has been used with great success to detect protozoan pathogens in humans (Burg et al., 1989; Lebech et al., 1992). The present critical situation of the oyster industry in the Gulf of Mexico, the southern Atlantic coast, and the Chesapeake Bay has strengthened the need for the development of sensitive and specific diagnostic assays for P. marinus to detect cryptic infec-
tions in oyster seed stocks, latent infections in overwintering oyster populations, the onset of infection in oyster larvae and spat, the presence of P. marinus in other marine organisms that may serve as secondary vectors or reservoirs, and the genetic structure of parasite field populations. In response to these needs, the present paper describes a semiquantitative PCR-based diagnostic assay for P. marinus DNA that provides both a rapid and sensitive in vivo assay for the detection of P. marinus in oyster hemolymph and gill tissues.
MATERIALS AND METHODS
Received 16 December1994;revised9 March1995;accepted9 March 1995. * To whom correspondenceshould be directed.
Total DNA was extractedfrom axenic culturesof P. marinususing a standardSDS/proteinase-Kprocedure(Ausubelet al., 1992). From a BamHl endonucleasedigestion,a 3.2-kb fragmentwas gel purifiedand cloned into the polylinkerof pBluescript(Stratagene,La Jolla, California). Both strandsof this clone were sequencedusing dideoxy terminators on an ABI automated DNA sequenceraccordingto the manufacturer's instructions. Sequence analysis using both GCG-FASTA searchesthroughGenBankand PAUP alignmentsrevealedthat the 3.2kb clone encoded the 5S and 18S rRNA genes separatedby a 1-kb noncoding domain. The development of a PCR-basedassay for this DNA fragmentfocused on the sequenceinformationof the noncoding domain between the two rRNA genes. Oligonucleotideprimerswere designedfor this regionusing the PRIMERprogram(V0.5, Whitehead Institute,Cambridge,Massachusetts)with stringentcriteria,including a requisitethat theirmeltingtemperaturesbe above 58 C. The best pair of primerswas the forwardsequence 5'-CACTTG TAT TGT GAA GCA CCC-3'and the reversesequence 5'-TTG GTG ACA TCT CCA AAT GAC-3', which would amplify a 307-bp target region (Fig. 1). These primerswere synthesizedon a BeckmanOligo1000 DNA synthesizer, quantifiedby optical density at 260 nm, and diluted to 100 gM workingstock solutions with sterile water. Oysterswere obtainedfrom 3 sources.One dozen oysterswere purchased from Mook Sea Farms, Damariscotta,Maine, to serve as negative (uninfected)controls.Fourteenoysterswereobtainedfrom 2 sites in Louisianaand shipped to us to serve as our primaryfield samples. We obtained 9 DNA samples that had been preparedfrom oyster gill tissues from individualscollectedat sites alongthe Gulf of Mexico and Atlanticseaboard.Inaddition,the DNA fromthe hemocytesofa heavily infected oyster from a previous study (stage 5 of the Mackin [1962] scale for the thioglycolateassay)was extractedfor use in this study as a positive infectioncontrol. A 1-mlsampleof hemolymphwasremovedfromthe adductormuscle
577
OFPARASITOLOGY, THEJOURNAL VOL.81, NO.4, AUGUST1995
578
5'
i
80 TGCTAGCCCATAGAACAGTCCAGTAGTTCAA CACTGTATTGTGAGCACCC 160 AACATTGTCCAAGGGGTGGAGGGGGATGCGCGAAATCGAT
240 AATTTCGCACATGGGCGAAATTGACTTGCAGGTGGGTATAAAAGTTGATGTAGGCCATGTGGCTCGATT GTCATTTGGAGTGTCACCAA GGTATGCTTCTGAGGATGGGGTGTTACAGTGGACCATATGAGGTA 3'
307
FIGURE1. Nucleotide sequenceof the 307-bp intergenicPerkinsusmarinusDNA domain that was targetedfor PCR amplificationin oyster DNA extracts.The locations of the oligonucleotideprimersare indicatedin the 5' and 3' boxes.
of each oysterthrougha notch in the shell. The hemocyteswerepelleted in a microcentrifugeand then processed by adaptinga spin-column methodologydesignedforthe rapidisolationof DNA fromhumanblood samples (Quiagen,Chatworth,California).The hemocytes were lysed in the presence of sodium dodecyl sulfate (SDS), proteinase-K,and guanidiniumHC1.The microscaleextractswere passed througha column matrix that binds double-strandedDNA (dsDNA) and washed severaltimes with 60%bufferedethanol to remove any contaminating proteinsand lipids. The DNA was eluted from the column with water in a volume of 20 gl. Sample optical density at 260 nm was used to quantifythe concentrationof DNA in eachextractandthe sampleswere then dilutedwith sterilewaterto a finalconcentrationof 1 g/,RlDNA. In order to set up a diagnosticPCR assay, each reactionhas to use a known amount of startingtemplate and there are several significant advantagesto adaptingthese separationcolumns to produceclean hemocyte DNA extracts.They do not requirethe use of organicsolvents (phenoland chloroform)that are requiredby standardextractiontechniques,which dramaticallyreducesthe handlingtime neededto prepare each sample.RNA is removedfromthe sampleso thata separateRNase digest is not requiredin orderto quantitatethe DNA on a spectrophotometer. The only limiting step in preparationtime is the numberof microcentrifugeslots available(with 2 centrifugeswe were able to process 24 samples in 30 min). All sampleswere subjectedto identicalreactionconditions for PCR amplificationin an EricompTwin-Block,water-cooledthermalcycler. A heat-stable Taq DNA polymerase was purchasedfrom Promega (Madison,Wisconsin)and each assay used 1.5 U of enzyme in a 25-al volume with the manufacturer'sreactionbuffer.In addition,each assay contained 1.5 mM MgCl2, 200 ,M each dNTP, 2 iM each primerand 1 gl (1 g) of template DNA. The temperatureprofilefor the amplification was 2' @94 C, 3' @61 C and 2' @72 C. This temperatureprofile was repeatedfor 35 cycles. Each PCR run startedwith a 5' @ 94 C denaturationand was completed with a 20' @ 72 C extension. Alternative procedureswere tested to includemore DNA polymerase,more amplificationcycles, higherand lower annealingtemperatures,higher primerconcentrations,and higherstartingtemplateconcentrations,but these did not increasethe assay's detection efficiency.The conditions listedabove weredeterminedto be the optimumreactioncharacteristics. PCR productswere resolved on a 2%agarosegel in the presenceof ethidiumbromide(EtBr;10 ng/ml finalconcentrationin gel) by loading 12.5 g1 of the 25-Mlreactionvolume into each well. A repetitive 123bp dsDNA size standard(Promega)was includedon the gels. Gels were photographedand then denaturedin 0.5 N NaOH with 1.5 M NaCl for 45', neutralizedin 1 M Tris-HCl(pH 7.2) with 1.5 M NaCl for 45', and blotted on nylon membranes (Schleicherand Schuell, Keene, New Hampshire)by capillarytransferovernight(Ausubelet al., 1992).DNA on the nylon membraneswas UV cross-linkedand the membranes stored dry at room temperature.Membraneswere prehybridizedfor severalhoursin 40%formamide,25 mM Na-PO4 (pH 7.2), 5 x standard saline citrates,0.1%SDS, 5 x Denhardt's,and 50 gg/ml yeast RNA at 42 C in a hybridizationoven. A PCR-amplifiedproductof the P. marinus 307-bp DNA targetdomain was radiolabeledby randompriming with a-32P-dCTP (3,000 Ci/mmol),addedto the hybridizationtubewith a fresh 10-ml aliquot of hybridizationbuffer(as above) and incubated overnightat 42 C. All PCRamplificationswerefirstresolvedon 2%agarosegelsto ensure that spuriousreactionproductswere not present.Followingthis visual inspection, 12.5-,l aliquots of each PCR amplificationwere directly
loaded onto nylon membranesusing a dot-blot apparatuswith gentle vacuum. The membraneswere then denaturedand neutralizedas describedfor the agarosegels in the above sectionand the DNA UV crosslinked.Hybridizationconditionsfollowedthe proceduredescribedabove for the Southernhybridizations. Kodak Biomax film was used for all radiographicexposuresbecause of the low backgroundinterferencefrom having emulsion on only one side of the film. The optimum lengthof time for exposingthe film was between 12 and 24 hr with intensifyingscreens at -80 C. For grain densitometry,autoradiographswere digitizedon a Microtekgray-scale scannerat 300 dpi and importedas TIFF files into Adobe Photoshop?. The histogramroutinein Photoshop? was used to estimatethe average pixel value (white = 0, black = 255) for a gel band or dot-blot, which is here reportedas autoradiographgraindensity. After hemolymphwas withdrawnfor DNA extraction,oysterswere processedfor determinationof P. marinusinfectionlevels basedon the fluidthioglycolateassayof Ray (1966). Rectaland mantletissuesamples (4 mm2)were placed in separatetubes containing 10 ml fluid thioglycolate medium (FTM), plus chloromycetinand mycostatin,and incubated for I wk in the dark at room temperature.Tissue smears were stained with Lugol's iodine solution and examined for a semiquantitative assessmentof infectionlevels based on a modifiedversion of the Mackin scale (Craiget al., 1989).
RESULTS Live oysters from 2 sites in Louisiana and 1 site in Maine, and oyster DNA from 9 sites around the Gulf of Mexico and Atlantic seaboard were screened for P. marinus infections with both the fluid thioglycolate assay and our PCR DNA assay. Five of the 8 individuals from ST3 in Louisiana were positive for P. marinus DNA with the PCR assay (Fig. 2A, C), whereas only 4 were positive for P. marinus hypnospores with the fluid thioglycolate assay (Table I). In this group, individual #2 did not produce a visible product on the EtBr-stained gel, but a hybridization signal was detected on the Southern blot (Fig. 2C). From the oyster DNA that had been obtained from gill tissues, 5 out of 9 individuals tested positive for P. marinus DNA (Fig. 2B). Because these samples were provided as gill-tissue DNA extracts, thioglycolate assays could not be carried out on these individuals. All PCR reactions were electrophoresed on 2% agarose gels and blotted for Southern hybridization autoradiography (Fig. 2D). None of the 6 individuals from ST 1 in Louisiana or the 10 individuals from Mook Sea Farms, Maine, tested positive for P. marinus with either assay (Table I). The amount of DNA extracted from a 1-ml hemolymph sample was variable, ranging from 10 to 3266tg total DNA (Table I). This level of variability necessitated standardizing the mass of DNA used in the PCR assays in order to accurately establish a detection limit and provide a normalization for quantitative estimates. All PCR amplifications used a standard amount of 1 ug total oyster DNA as a template. The samples from both
INOYSTERS 579 ETAL.-PCR ASSAYFORP. MARINUS MARSH
I
C01NON I
II
I
I
I
I
369> 246 > 123 >
307>
< 307
4< 307
FIGURE2. Amplificationof the PerkinsusmarinusDNA targetin DNA extractsfrom fieldcollectedoysters.A: Eightoysterhemocytesamples of the agarose from ST3, Louisiana.B: Nine oyster-gillDNA samplesfrom Louisianato SouthCarolina.C and D: Southernblot autoradiography gels picturedin A and B, respectively.The numbersto the left of panel A indicatethe dsDNA size standardsin numbersof bp; the 307-bp target band position is indicatedadjacentto the remainingpanels.
Louisiana sites were run a second time using 10 gg of total oyster DNA as template in a PCR amplification. This increase in template concentration did not reveal any positive P. marinus infections that had not been detected with the 1-gg DNA assays; however, the increase in template concentration resulted in the appearance of large (> 1-kb), nonspecific amplification products, but none of these secondary bands hybridized with the P. marinus DNA probe in Southern blots of the agarose resolving gels. Thus, 1 gg of starting template DNA was judged to be appropriate for maximizing the detection sensitivity and minimizing random priming events during the PCR reaction. The relationship between amplification intensity and starting template concentration in the PCR amplification was assessed by a spike-recovery assay that consisted of 2, 10-fold serialdilution series of P. marinus total DNA, using either water or water with oyster DNA at a concentration of 1 gg/gl. A 1-gl aliquot of each dilution was then used as template in a PCR amplification of P. marinus DNA under the same condition as the unknown oyster samples. The amplified products were electrophoresed on a 2% agarose gel, and under EtBr staining a visible product was apparent at 10 pg of starting P. marinus total DNA (Fig. 3A). With Southern hybridizations of these gels, a positive amplification product was evident down to as little as 100 fg of starting P. marinus total DNA (Fig. 3B), and with dot-blot hybridizations a signal could be detected with 10 fg of starting P. marinus total DNA (Fig. 3C). Comparing EtBr visualization to dot-blot hybridization represents an increase in assay sensitivity spanning 6 orders of magnitude. There was no difference in the sensitivity or specificity of assays run with or without the presence of oyster DNA, indicating that in the unknown samples, the oyster DNA does not interact with the primers and interfere with the assay's amplification efficiency. The quantitative relationship between P. marinus DNA target amplification and initial P. marinus DNA concentration was
measured by autoradiographic grain density counts of both the hybridization bands in Southern blots of the agarose gels and direct dot-blots of the PCR products (Fig. 4). Both sets of measurements evidence typical sigmoidal saturation curves. The dot-blot measurements show a greater sensitivity level (10 fg) than the Southern blots, but a steeper saturation slope. There is an obvious relationship between the starting template concentration and final amplification of a P. marinus DNA target even in the presence of a much larger amount of oyster DNA (1 Mg,or up to 8 orders of magnitude more than the P. marinus detection level). The results presented in Figure 4 indicate that it is possible to semiquantify the amount of P. marinus DNA in a sample by estimating the lowest dilution level that is necessary to extinguish any amplification of target. Because there was no detectable difference either with or without the presence of oyster DNA in the standard-curve assays, the samples that were scored positive for P. marinus infection in Figure 2 were serially diluted 10-fold with water. A 1-Mlaliquot of each dilution was then used as template in PCR amplifications. Reaction products were dot-blot hybridized and grain counts were made of the autoradiographs. From these data, a dilution endpoint titer was determined for each sample as the template dilution level at which the dot-blot hybridization signal could no longer be distinguished from the background signal. These values ranged from 1:10' for Louisiana ST3 individual #2, to 1:106for a heavily infected oyster (OTG, Table I) that was used in these studies as a positive infection control. By assigning a value of 1 to the dilution level at which the amplification signal was extinguished, a titer for P. marinus DNA could be estimated for each preceding dilution, and the titer curves for the unknown samples evidenced similar sigmoidal saturation kinetics as the standards in Figure 4. The dilution endpoint PCR amplifications thus provide a semiquantitative estimate (to the nearest power of 10 in this
580
THEJOURNAL OFPARASITOLOGY, VOL.81, NO.4, AUGUST1995
TABLEI. Fluid thioglycollate assay for Perkinsus marinus hypnospores using the Mackin scale with the corresponding DNA levels and PCR screening results for P. marinus DNA in each sample.* Tissue sample Oyster ID
Rectal
PCR
0 0 0 0 0 0 0 0
30 10 26 106 41 23 116 25
0 0 0 0 0 0 0 0
0 0
0 0
117 100
0 0
Station 1, Louisiana STI-1 0 ST1-2 0 0 ST1-3
0 0 0
183 40 211
0 0 0
MK-9 MK-10
ST1-4
0
0
16
0
ST1-5 ST1-6
0 0
0 0
188 246
0 0
2.00 0 2.00 0 0 0
115 236
102 10,
195
103
102 89 66
Station 3, Louisiana ST3-1 4.33 ST3-2 0 ST3-3 3.33 0 ST3-4 0 ST3-5 ST3-6 0
o o-00
pg
fg 0
O-
DNA (Mg)
Mook Sea Farms, Maine MK-1 0 MK-2 0 MK-3 0 MK-4 0 MK-5 0 MK-6 0 MK-7 0 MK-8 0
Mantle
ng
ST3-7
5.00
4.33
85
0 0 0 104
ST3-8
1.00
1.00
326
104
307>
307>
(
ng pg 10 1 100 10
fg 1 100 10
1
Infected control
3. Amplificationof the PerkinsusmarinusDNA targetusing FIGURE a knownamountof total P. marinusDNA in a 10 x serialdilutionwith * The total mass of DNA extractedfrom a 1-ml hemolymphsample from each a constantlevel of oystergenomic DNA (1 tg/tl). Valuesacrossthe top oysteris presentedin Ag;the columnlabeledPCRpresentsthe dilutionendpoint of the panels representthe total mass of P. marinus DNA that was titersforP. marinusDNA thatwereestimatedin the sampleswith a 0 indicating added to each reaction as template. A: EtBr visualizationof agarose that no infectionwas detected. resolvinggel, 10 pg detection limit (with 12.5 Al of 25 Al PCR amplificationloaded in each lane). B: Southernblot of above agarosegel, 100 fg detection limit. C: Dot-blot hybridizationof PCR amplification,10 fg detection limit (with 12.5 Al of 25 Al PCR amplificationloaded in case) of the initial concentration of P. marinus DNA in oyster each well). The 307-bp targetsize is indicatedin panels A and B. hemolymph extracts. At present, we are unable to convert this mass amount of P. marinus DNA into an estimate of genome copy number and thus translate our results into an estimated fections is the lack of a sensitive assay that would allow for both P. marinus cell count. The in vitro cultures of P. marinus are the detection of P. marinus at low infection levels and the disdominated by cell groups that do not dissociate after dividing, crimination between putative geographic subpopulations of P. making it difficult to get accurate estimates of cell numbers in marinus. The fluid thioglycolate assay (FTM) is currently the a reference sample. The PCR estimates of initial P. marinus most commonly used diagnostic test to identify P. marinus (Ray, DNA in each sample were compared to the hybridization in1966). In this medium, P. marinus trophozoites are stimulated tensity of the Southern blots of the initial PCR amplifications to develop into hypnospores. After staining with Lugol's iodine (from Fig. 2) and the fluid thioglycolate stage values to determine these hypnospores can be visualized by microscopic solution, the degree of coherence between these 2 techniques (Fig. 5). examination under a low-power objective. The primary experPCR gel band intensity is linearly related to the estimate of P. imental limitations of this technique are a high variance commarinus DNA titer (r2 = 0.902), whereas the correlation between associated with measuring low-level infections (Ray, ponent P. marinus DNA titer and thioglycolate stage is much lower. 1952, 1966; Choi et al., 1989; Gauthier and Fisher, 1990; Bushek et al., 1994) and a limitation in sensitivity because the only DISCUSSION life stage detected are trophozoites that are induced to undergo The most significant obstacle to developing effective treat- hypnospore development (Ragone and Burreson, 1993). The ment and management strategies for controlling P. marinus in- low cost, ease of application, and overall efficacy of the FTM OGT
5.00
4.33
15
106
INOYSTERS 581 ETAL.-PCR ASSAYFORP. MARINUS MARSH
160 (D 140 C
/
120
/
L100
80U 60 M 0.
40o
0o 20
3
EndPoint
/
*
Dot
Blot
oEndPinto Southern Blot
/
1fg
10fg 100fg lpg
Template
DNA
in
10pg 100pg ing Dilution
End-Point
10ng
PCR
FIGURE4. Standardcurves for the dot-blot and Southernblot quantificationof the amplifiedPerkinsusmarinusDNA targetas a functionof the amount of total P. marinusDNA that was used in the amplificationreaction.All reactionswere run with 1 Agof oyster DNA present.The dilution endpointof both samplesis indicatedas the last dilution in which an amplifiedPCR productcould be detectedwith eithermethodology. Grain density representsthe digitized median pixel value within the band or dot area (0 = white; 255 = black).
assay will ensure its continued use as a diagnostic test for routine screening of P. marinus infection levels in field populations of oysters. However, at present there is a need for developing a diagnostic assay that is: (1) sensitive enough to detect the presence of any P. marinus life stage at low levels in either overwintering populations, spat, or juveniles; and (2) robust enough to discriminate between putative geographic races or strains of P. marinus. Antibody-based assays for the detection of P. marinus proteins in oyster tissues have recently been used with mixed success at meeting the first requirement above (Choi et al., 1991; Dungan and Roberson, 1993). However, Fong et al. (1993) recognized the potential for nucleic acid-based assays to fulfill both methodological requirements because of their sensitivity, specificity, and ability to provide genetic sequence information. The largest limitation to using DNA targets for diagnostic assays is the degree of sequence identity that can exist in homologous genes between a parasite and its host. Fong et al. (1993) report the PCR cloning and sequencing of the small subunit (ssu) rRNA gene from P. marinus; however, there was a 77% sequence identity between the P. marinus ssu-rRNA gene and the oyster ssurRNA. The ideal target domains for PCR-based diagnostic assays are thus noncoding regions where there is a greater probability of sequence divergence between host and parasite. Our work with total DNA from axenic cultures of P. marinus has yielded a 3.2-kb DNA clone with an intergenic domain
between the 5S and 17S rRNA genes. Oligonucleotides designed to this region are functionally specific for P. marinus DNA under stringent annealing conditions. This specificity is the basis for the sensitivity in screening oysters for the diagnosis of P. marinus infections. Our assays were standardized by using 1 Ig of total oyster DNA, and we have shown that it is possible to amplify a detectable product with 108 (8 orders of magnitude) less P. marinus total DNA. Although the oligonucleotides used in the PCR assay were highly specific for P. marinus, in some of the PCR amplifications, random products were produced in the 1.0-1.5-kb size range, especially when 10 ,g of oyster DNA was used as template in a PCR amplification. The origin of these products is unknown, but in no instances did any of these spurious bands evidence a positive hybridization signal with the P. marinus DNA probe in Southern blots of the agarose gels. There is a possibility that these bands are not derived from mispriming on oyster DNA because all reactions were standardized with 1 ,ggof oyster DNA as template; the appearance of spurious bands was highly variable between sample groups and primarily limited to the individuals from Mook Sea Farms, Maine. The oligonucleotide primers may be recognizing DNA from other protozoans with differences in microflora between samples then accounting for the variance in the pattern of these amplified secondary bands. We are currently evaluating other intergenic domains to determine if these products can be completely eliminated.
582
VOL.81, NO.4, AUGUST1995 OFPARASITOLOGY, THEJOURNAL
0 n140-
C
- 6
OTGA
Southern
20
0
120 O S0
co
10
oo.1080
10
10 o
c
0
c
-4
-
60
yi
Y
0 4
10
10
+ 57.219 2
11.3980ogx
2
-
40
•
20-
13
1012 00
Dilution
End-Point
14
15
16
(D
Titer
of the amplified FIGURE5. Infectionintensityin oysterhemocytesas determinedby the thioglycolateassayand Southernblot autoradiography PerkinsusmarinusDNA target.Both variablesare plottedagainstthe level of P. marinusDNA in each sampleas estimatedby dilutionendpoint PCR using a dot-blot hybridizationto identify amplificationextinction. A linear regressionis presentedfor Southernblot band density as a function of dilution endpoint titer (y = 11.398[log(x)]+ 57.219; r2 = 0.902; n = 8). Grain density representsthe digitizedmedian pixel value within the band or dot area (0 = white; 255 = black).
Any quantitative diagnostic assay requires a rigorously established detection limit. The only way to conclude positively that a negative assay result is indicative of the absence of an infection below this detection limit is to standardize all diagnostic assays performed. Although our presentation of a PCRassay does not present a large enough sample size for a rigorously defined detection limit, our use of a standardized amount of oyster DNA in each amplification can provide us with an estimate of this detection limit. For the PCR assay, 1 Ig of total oyster DNA establishes a normalization so we can confidently conclude that a negative result means that there is less than 1 fg P. marinus DNA per 1 gg of oyster DNA. The high variability in the oyster hemolymph DNA values in Table I is important because it illustrates that a diagnostic assay cannot be based simply on a fixed volume of hemolymph, as is the case with the FTM assay. From the agarose gels and Southern blots in Figure 3, it is apparent that there are distinct differences in the amplification intensity of the P. marinus DNA target. The most likely source of these differences is the amount of P. marinus DNA in each of the oyster sample DNA extracts. Most quantitative PCR strategies essentially involve some form of a competitive assay in which the amplification of a known template is employed to calculate an efficiency that is subsequently used to convert the amplification of an unknown back to its starting template concentration (see Innis et al., 1990). These techniques all require a genetically engineered "standard" target and a thorough quantification of reaction kinetics.
In contrast, we have developed a semiquantitative assay that can be performed on any sample without any prior preparation or standardization. It is based on identifying the lowest dilution at which the amplification of a specific target sequence is no longer detectable. Limiting dilution assays are routinely used for many cell biology applications, but only recently have such assays been developed for the detection sensitivity of PCR (Sykes et al., 1992). The accuracy of such an assay is only as fine as the dilution level employed to titrate the endpoint, but the precision in our samples is high and there appears to be no effect by the presence of significantly higher levels of oyster DNA. We refer to this technique as dilution endpoint PCR. Furthermore, a significant linear regression correlates band grain density on a Southern blot to the estimated dilution endpoint titer of the sample (Fig. 5). The implication of this relationship is that it may not be necessary to perform a dilution-series assay on every sample to be quantified. It may be possible to establish a standard regression and from the band grain density of a Southern autoradiograph, then to estimate directly the relative P. marinus DNA concentration. Estimating an infection level to the nearest power of 10 may not appear to be an accurate measure, but it may provide the degree of quantification necessary to determine changes in oyster infection levels in response to experimental manipulations. Our preliminary comparisons of FTM and PCR assay methodologies in this paper have use different tissues (rectal and mantle tissues in the former, and hemocyte and gill DNA in the latter). However, a linear relationship between the Mackin
MARSH ETAL.-PCR ASSAYFORP. MARINUS INOYSTERS 583
scale for hemocyte infection levels and systemic infection levels in mantle tissues has been demonstrated (Gauthier and Fisher, 1990). Although P. marinus infection levels may vary between tissues (Choi et al., 1989), it is believed that hemocytes serve as one of the primary vehicles by which the pathogen is translocated throughout the body of an oyster host, whereas rectal and mantle tissues serve as incubation sites once the disease is established. Thus, our testing of a PCR-based assay focused on using hemocytes because of the sensitivity required for detecting the onset of an infection, whereas the FTM validation used rectal and mantle tissues because of the high incidence of P. marinus in these tissues once an infection has been established. This is likely the situation for the ST3-2 individual that was positive for an infection with the PCR assay of hemocytes but negative for an infection with the FTM assay of rectal and mantle tissues. The low dilution endpoint titer estimate of 10' in this individual suggests that an infection may not have been fully established. Future studies and optimization of PCR-based assays will undoubtedly have to consider different tissue sources. Our first investigation has focused on hemocytes primarily because they represent a cell type that can be sampled nondestructively and from which it is convenient to extract total DNA. These 2 characteristics offer advantages for the routine screening of oysters for low infection levels. In summary, we have developed and evaluated a PCR-based diagnostic assay for the detection and quantification ofP. marinus DNA in oyster DNA extracts. This technique provides a rapid and reliable assessment of P. marinus infection levels but is not designed to replace the widely used FTM assay. The PCRbased assay requires a facility equipped and supplied for the analysis of nucleic acids and as such cannot be instantly incorporated into the routine operations of all the laboratories currently active in P. marinus research. The amount of sample preparation and the expense of supplies for the PCR assay make this an unlikely candidate to replace the FTM assay for the routine monitoring of infection levels in field populations. However, the PCR assay establishes a new diagnostic procedure that provides a level of sensitivity and quantification that is not afforded by the FTM assay. Consequently, it will open new avenues of experimental research regarding the etiology of this oyster disease by allowing for the detection and enumeration of P. marinus in either cryptic infections, different oyster life stages, secondary hosts/vectors, or experimental applications to assess factors that may alter parasite virulence.
ACKNOWLEDGMENTS We gratefully acknowledge the contributions of Thomas Soniat for providing oysters from his study sites in Louisiana and of Matt Hare for providing the DNA samples that were prepared from oysters collected at several sites in the southeastern U.S.A. This work was supported by DOC Cooperative Agreement NA47FL-0163 NOAA/NMFS Oyster Disease Research Program to G.R.V.
LITERATURE CITED J. D. 1988. Epizootiologyof the diseasecausedby the oyster ANDREws, pathogenPerkinsusmarinusand its effectson the oyster industry.
In Disease processesin marinebivalve molluscs,W. S. Fisher(ed.). AmericanFisheriesSociety SpecialPublicationVol. 18, Academic Press, New York, New York, p. 281-295. F. M., R. BRENT, R. E. KINGSTON, D. D. MOORE, J. G. SEIDAUSUBEL, ANDK. STRUHL.1992. Shortprotocolsin moMAN,J. A. SMITH, lecular biology, 2nd ed. John Wiley and Sons, New York, New York, 685 p.
BURG, J., C. GROVER, P. POULETTY,ANDJ. BOOTHROYD. 1989. Direct
and sensitive detection of a pathogenicprotozoan, Toxoplasma. Journalof Clinical Microbiology27: 1787-1792. BUSHEK, D., S. E. FORD,ANDS. K. ALLEN.1994. Evaluationof methods using Ray's fluid thioglycollatemedium for diagnosisof Perkinsusmarinusinfectionin the easternoyster,Crassostreavirginica. Annual Review of Fish Diseases 4: 201-217. P. F. FRELIER, CHOI,K. S., D. H. LEWIs,E. N. POWELL, ANDS. M. RAY.
1991. A polyclonal antibody developed from Perkinsusmarinus hypnosporesfails to cross reactwith otherlife stagesof P. marinus in oyster (Crassostreavirginica)tissues. Journalof ShellfishResearch10: 411-415. E. A. WILSON, D. H. LEWIs,E. N. POWELL, ANDS. M. RAY. -, 1989. The energeticcost of Perkinsusmarinusparasitismin oysters: Quantificationof the thioglycollatemethod. Journalof Shellfish Research8: 125-131. R. R. FAY,ANDJ. M. BROOKS.1989. DisCRAIG, A., E. N. POWELL, tribution of Perkinsusmarinusin Gulf Coast oyster populations. Estuaries12: 82-91. DUNGAN,C. F., AND B. S. ROBERSON.1993. Binding specificities of
mono- and polyclonalantibodiesto the protozoanoysterpathogen Perkinsusmarinus.Diseases of AquaticOrganisms15: 9-22.
K. Koo, J. SUN, M. SOGIN,D. BUSHEK,D. FONG,D., R. RODRIGUEZ, T. L. LITTLEWOOD,ANDS. FORD. 1993. Small subunit ribosomal
RNA gene sequenceof the oysterparasitePerkinsusmarinus.MolecularMarineBiology and Biotechnology2: 346-350.
GAUTHIER,J. D., AND W. S. FISHER. 1990. Hemolymph assay for
diagnosis of Perkinsusmarinusdisease in oysters Crassostreavirginica (Gmelin, 1791). Journalof ShellfishResearch9: 367-371. ANDT. WHITE.1990. PCR proJ. SNINSKY, INNIS,M., D. GELFAND, tocols: A guide to methods and applications.AcademicPressInc., New York, New York, 482 p. LEBECH, M., A. LEBECH, S. NELSING, J. VUUsT, L. MATHIESEN, AND E. PETERSEN. 1992. Detection of Toxoplasmagondii DNA by poly-
merase chain reaction in cerebrospinalfluid from AIDS patients with cerebraltoxoplasmosis. Journalof Infectious Diseases 165: 982-983. LEVINE,N. D. 1978. Perkinsusand other new taxa in the protozoan phylum Apicomplexa.Journalof Parasitology64: 549. J. G. 1962. Oysterdisease caused by DermocystidiummarMACKIN, inum and other microorganismsin Louisiana.Publicationof the Instituteof MarineSciences,University of Texas 7: 132-229. F. 0. 1966. Dermocystidiummarinuminfection in oysters. PERKINS, MarineFisheriesReview 38: 19-21. 1988. Structureof protistanparasitesfound in bivalve molluscs.In Disease processesin marinebivalve molluscs,W. S. Fisher (ed.). AmericanFisheriesSociety SpecialPublicationVol. 18, Academic Press, New York, New York, p. 93-111. L. M., ANDE. M. BURRESON. 1993. Effect of salinity on RAGONE, infection and pathogenicityof Perkinsus marinus in the eastern
oyster, Crassostrea virginica. Journal of Shellfish Research 12: 1-7. RAY, S. M. 1952. A culturetechniquefor the diagnosis of infections
with DermocystidiummarinumMackin,Owen and Collierin oysters. Science 116: 360-361. 1966. A review of the culturemethod for detectingDermocystidiummarinum,with suggestedmodificationsand precautions. Proceedingsof the National ShellfishAssociation54: 55-69.
SYKES,P. J., S. H. NEOH, M. J. BRISCO,E. HUGHES,J. CONDON,AND
A. A. MORLEY.1992. Quantitationof targetsfor PCR by use of limiting dilution. Biotechniques13: 444-449.
