Rapid polyvalent screening for largescale environmental Spiroplasma surveys.

Surface serology is an important determinant in Spiroplasma systematics. Reciprocal antigen/antibody reactions between spiroplasmas and individual antisera delineate the 38 described groups and species. However, reciprocal serology is impractical for large-scale studies. This report describes a successful, streamlined polyvalent screening approach used to examine isolates from an environmental survey.

The spiroplasma deformation (DF) test is a central feature of spiroplasma systematics that examines the cell surface antigenicity of clonal isolates. Originally adopted for detecting Spiroplasma poulsonii (then a “noncultivable” agent causing sex ratio abnormalities in Drosophila willistoni), the technique was modified for general use in spiroplasma systematics (15,16). The DF test has subsequently been used, in part, to define spiroplasma groups (6,12,17) and species (1). There are currently 38 described serogroups (reviewed in 7; 14), and some of the groups are further subdivided. For example, group I has eight subgroups that include five characterized species (7). The serogroup I subgroups have 0.986-0.991 16S rDNA sequence similarity, yet distinctive serologies (5).

Horse flies or tabanids (Diptera:Tabanidae) are a rich source of spiroplasmas (2,4,13). A large-scale biodiversity survey of tabanid-associated spiroplasmas began in the southeastern United States in 1987, with most spiroplasmas being isolated from female tabanids in southeast Georgia (Bulloch County). Based on DF tests, nine groups and five subgroups (of group VIII) were isolated in this initial survey. When the study was extended to tabanid-associated spiroplasmas in Costa Rica, Ecuador and Australia, the number of putative groups increased even more rapidly than in the southeastern United States. To date, over 1,000 spiroplasma isolations have been completed from tabanid flies as part of this environmental survey.

Faced with a deluge of new isolates, the standard DF test that individually examines the reciprocal serology of each antigen/antiserum pair was impractical. Instead, a polyvalent screen (PVS) was adopted that used mixtures of antisera directed against spiroplasma strains known to occur in tabanids. The initial PVS trial screened more than 400 isolates from the southeastern United States and the Costa Rican highlands using a total of 13 antisera (all against known U.S.A. tabanid-associated strains) distributed in 6 screening cocktails (Table 1). As the study progressed, the number and content of the screening mixtures were modified. For example, the screening array was expanded to 24 antisera in 12 cocktails to process isolates from the Costa Rican lowlands (Table 1), and later it was expanded further to include 38 antisera in 12 cocktails to screen Australian and Ecuadorian isolates (Table 1). Initially, the components of each cocktail were selected based on serological relationships. For example, one antisera cocktail (Table 1, cocktail 2) contained all group VIII strains that had low level cross-reactivity (8). As the study progressed and novel serogroups were identified, the new antisera were included in the cocktails with the fewest number of component antisera. This approach was dynamic and efficient, as antisera against new serogroups/strains could be added to screening cocktails as needed.

Table 1

Polyvalent screen antisera cocktails for tabanid-associated spiroplasmas.

Antisera Mixture No.	First Screen: U.S.A. & Costa Rica Highlands			Second Screen : Costa Rica Lowlands			Third Screen: Australia & Ecuador
	Group No.	Species	Strain	Group No.	Species	Strain	Group No.	Species	Strain
1	IV	S. apis	B 31	IV	S. apis	B 31	IV	S. apis	B 31
	IV	S. sp.	PPS-1	IV	S. sp.	PPS-1	IV	S. sp.	PPS-1
							IV	S. sp.	W-13
2	VIII-1	S. syrphidicola	EA-1	VIII-1	S. syrphidicola	EA-1	VIII-1	S. syrphidicola	EA-1
	VIII-2	S. chrysopicola	DF-1	VIII-2	S. chrysopicola	DF-1	VIII-2	S. chrysopicola	DF-1
	VIII-3	S. sp.	TAAS-1	VIII-3	S. sp.	TAAS-1	VIII-3	S. sp.	TAAS-1
				VIII	S. sp.	BARC 2649	VIII	S. sp.	BARC 2649
				VIII	S. sp.	BARC 1357	VIII	S. sp.	BARC 1357
3	XIV	S. corruscae	EC-1	XIV	S. corruscae	EC-1	XIV	S. corruscae	EC-1
	XXXI	S. montanense	HYOS-1	XXXIV	S. sp.	BARC 1901	XXXIV	S. sp.	BARC 1901
4	XXIII	S. gladiatoris	TG-1	XVII	S. turonicum	Tab4C	XVII	S. turonicum	Tab4C
	XXXIII	S. tabanidicola	TAUS-1	XXXI	S. montanense	HYOS-1	XXXI	S. montanense	HYOS-1
5	XVIII	S. litorale	TN-1	XVIII	S. litorale	TN-1	XVIII	S. litorale	TN-1
	XXXII	S. helicoides	TABS-2				XVIII	S. sp.	TAAS-2
							NG1	S. sp.	GSU5529
							NG	S. sp.	GSU5603
6	XXXIV	S. sp.	BARC 1901	XXIII	S. gladiatoris	TG-1	XXIII	S. gladiatoris	TG-1
	VIII	S. sp.	BARC 2649	XXIII	S. sp.	BARC 4689	XXIII	S. sp.	BARC 4689
							XXIII	S. sp.	TASS-1
7				XXXII	S. helicoides	TABS-2	XXXII	S. helicoides	TABS-2
							NG	S. sp.	GSU5478
8				XXXIII	S. tabanidicola	TAUS-1	XXXIII	S. tabanidicola	TAUS-1
				XXXIII	S. sp.	TABS-1	XXXIII	S. sp.	TABS-1
							NG	S. sp.	GSU5366
9				XXXVII	S. sp.	BARC 4908	XXXVII	S. sp.	BARC 4908
				XXXVII	S. sp.	BARC 4906	XXXVII	S. sp.	BARC 4906
				XXXVII	S. sp.	BARC 4907	XXXVII	S. sp.	BARC 4907
							NG	S. sp.	GSU5420
							NG	S. sp.	GSU5373
10				XXXV	S. sp.	BARC 4886	XXXV	S. sp.	BARC 4886
11				XXXVI	S. sp.	BARC 4900	XXXVI	S. sp.	BARC 4900
							NG	S. sp.	GSU5360
							NG	S. sp.	GSU5400
12				XXVII	S. lineolae	TALS-2	XXVII	S. lineolae	TALS-2
				XXVII	S. sp.	BARC 4903	XXVII	S. sp.	BARC 4903
							XXXVIII	S. sp.	GSU5450
							NG	S. sp.	GSU5446H

No group designation.

The screening cocktails were generated by combining equal volumes of each antiserum in M1D medium (10) to achieve a final dilution of 1:10 for each. All antisera were combined with other sera successfully, in that all cocktails reacted strongly when challenged with an antigen homologous to a component of the cocktail. These cocktails were routinely stored at 4°C for up to one year with no detectable loss in reactivity. Isolates were screened by mixing an equal volume of actively growing spiroplasma culture with each antisera cocktail in microtiter plate wells (20 µl each), and plates were incubated at room temperature for 30 minutes. Samples from each of the PVS reactions were examined by dark-field microscopy at 1200X for cell deformation (Figure 1). Reaction mixtures that exhibited >50% cell deformation (e.g. clumping, blebs) were considered positive for antigenic reactivity to a component of the cocktail.

Figure 1

Deformation (DF) test with Spiroplasma citri strain Maroc R8A2 and homologous antiserum with an endpoint DF titer of 1:2048. Cultures were incubated for 2 hours at room temperature without antiserum (A) as a control for normal morphology or with varying dilutions of antiserum exhibiting intermediate (B) to strong (C) deformation reactions. For image generation, cells were fixed overnight in 1.5% glutaraldehyde, resuspended in fresh M1D medium, and then photographed under dark-field illumination.

The PVS rapidly revealed that most of the new isolates were matches to known serogroups. All isolates that were positive to a screening cocktail were tested in individual DF tests using each of the component antisera; the two-fold dilution series examined antisera titers of 1:20 to 1:2560. Generally, an isolate reacting with antiserum to a described group at a concentration <1:320 was considered to be a member of that group. An isolate exhibiting no or only low levels of cross reactivity (1:20-1:40) to all antisera in the PVS was tentatively given candidate status as a new serovar. When the original isolations were mixed cultures or when adaptation to laboratory media was required for optimal morphology (14), additional passages and/or limiting dilution cloning (3,11) were necessary before serological placement could be achieved.

As new groups emerged, antisera to triply-cloned strains were prepared and added to the PVS and DF tests. For the described study, 23 clusters of isolates with novel serology (i.e. putative new serogroups) were identified. Antisera production was completed as described previously (8,9), with the substitution of MPL®+TDM, MPL®+TDM+CWS or TiterMax® Gold (Sigma-Aldrich Chemical Company, St Louis, Missouri, U.S.A.) as the adjuvant. No booster injection was required when TiterMax® Gold was used. Sera were aliquoted and stored at -70°C or lyophilized. Animals were cared for in accordance with approved guidelines set forth in the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 86-23), and their use was reviewed and approved by the Institutional Animal Care and Use Committee at Georgia Southern University.

With hundreds of spiroplasma isolates from tabanid flies, it was prudent to develop an efficient preliminary screening system. The PVS procedure permitted initial screening of new isolates with 6-12 antisera mixtures rather than 13-38 individual antisera, streamlining the process and reducing the amount of antisera required. For example, the ten field isolates from Ecuador required 27 PVS and 150 DF tests to resolve five new groups and one geographic variant. The same serological analyses without the PVS test would have required an estimated 1,178 DF tests. Inclusion of the PVS test reduced the amount of antisera required by 44% and the time/labor commitment by approximately 85%. As the number of new serogroups and species continues to increase, the need for streamlined serological screening becomes even more critical.

For large environmental surveys we suggest the following screening procedures: (A) complete a PVS for early isolate passages using antisera combinations of strains that are likely to be related (e.g. common host); (B) complete one to two rounds of dilution cloning for non-reactive isolates (i.e. those showing <50% deformation) and then repeat the PVS; (C) triply clone non-reactive isolates, repeat PVS, complete DF tests versus all strains in cocktails that showed any level of cross-reactivity in the final PVS, and then produce antisera against any unresolved isolates (i.e. no reactivity at concentrations <1:320); and (D) use 16S rRNA-based phylogenetic analyses to identify closely related strains and then DF test the isolate versus all closely related strains or groups not included in the PVS. As an example, screening data for environmental isolates from the Costa Rican highlands is shown in Table 2; this approach successfully identified 4 putative new species (represented by BARC 4886, BARC 4900, BARC 4908/4907/4906 and GSU5450); serogroup VIII isolates (BARC 4898/4899); and geographic variants (BARC 4901/4902/4903/4905) of S. lineolae (14). Further characterization is required for serogroup (12) or species (1) designations.

Table 2

Resolution matrix for 13 spiroplasma isolates from Poeciloderas quadripunctatus tabanid flies in the Costa Rican highlands using polyvalent screening (PVS) mixtures 1-6.

Isolate	PVS #12	PVS #23	PVS #34	Individual one-way DF (titer) 5	Antisera produced (titer)6	Reactive strain(s) in reciprocal DF tests (titer) 7	Reactive antisera in reciprocal DF tests (titer) 7	Serogroup Placement8
BARC 4886	+ (5)	NA9	+ (5)	S. helicoides (≤80)	Y (640)	none	S. helicoides (≤80)	XXXV10
BARC 4898	+ (2)	NA	+ (2)	TAAS-1 (≥640)	N	NA	NA	VIII
BARC 4899	+ (2)	NA	+ (2)	TAAS-1 (≥640)	N	NA	NA	VIII
BARC 4900	none	none	none	none	Y (2560)	none	none	XXXVI10
BARC 4901	none	none	none	S. lineolae11 (320)	N	NA	NA	XXVII
BARC 4902	none	none	none	S. lineolae (320)	N	NA	NA	XXVII
BARC 4903	none	none	none	S. lineolae (1280)	Y (2560)	S. lineolae (1280)	S. lineolae (320)	XXVII
BARC 4904	none	none	none	BARC 4900 (1280)	N	NA	NA	XXXVI10
BARC 4905	none	none	none	S. lineolae (320)	N	NA	NA	XXVII
BARC 4906	+ (5)	NA	+ (5)	S. helicoides (80)	Y (640)	BARC 4907 (160)	BARC 4907 (160)	XXXVII10
						BARC 4908 (40)	BARC 4908 (40)
							S. helicoides (80)
BARC 4907	+ (5)	NA	+ (5)	S. helicoides (40)	Y (320)	BARC 4906 (160)	BARC 4906 (160)	XXXVII10
						BARC 4908 (160)	BARC 4908 (160)
BARC 4908	+ (5)	NA	+ (5)	S. helicoides (40)	Y (640)	BARC 4906 (40)	BARC 4906 (40)	XXXVII10
						BARC 4907 (160)	BARC 4907 (160)
GSU5450	+ (5)	NA	+ (5)	S. helicoides (40)	Y (1280)	none	S. helicoides (40)	XXXVIII10

Polyvalent screening mixtures contained antisera prepared against the following strains: Mixture 1 (B31 [S. apis], PPS-1); Mixture 2 (EA-1 [S. syrphidicola], DF-1 [S. chrysopicola], TAAS-1); Mixture 3 (EC-1 [S. corruscae], HYOS-1 [S. montanense]); Mixture 4 (TG-1 [S. gladiatoris], TAUS-1 [S. tabanidicola]); Mixture 5 (TN-1 [S. litorale], TABS-2 [S. helicoides]); Mixture 6 (BARC 1901, BARC 2649). The six Costa Rican highland serogroups were previously described (14), but that report did not provide the detailed resolution matrix data.

Early passage isolates were screened with antisera mixtures; mixtures that resulted in positive reactions (≥50% deformation) are indicated in parentheses.

Nonreactive isolates were subjected to 1-2 rounds of dilution cloning and then retested with the PVS mixtures.

All isolates were triply cloned and then retested with PVS mixtures.

One-way deformation (DF) tests were completed using individual antisera from any mixture that showed any level of reactivity in the PVS.

Antisera were produced for nonreactive and representative isolates; titers against antigenic homologues are indicated.

Strains or antisera are listed for reactions at titers ≥40.

One-way DF titers ≥320 allowed tentative serogroup placement; tentative isolate placement in a previously described serogroup generally served as the testing end-point.

Not applicable based on test procedures or prior screening results.

Member of a novel serogroup identified in the Costa Rican highlands (14).

S. lineolae was not included in the original PVS screen but was tested individually against all isolates as a closely related species.

Inclusion of a PVS prior to DF tests makes large-scale environmental surveys to examine spiroplasma biodiversity and biogeography feasible, and the approach may have wider applications for other mollicutes that also rely on serology for classification.