点击显示 收起
Laboratory of Parasitic Diseases, National Institutes of Health, Bethesda, Maryland
Received 9 May 2002; revised 29 August 2002; electronically published 13 December 2002.
The clinical manifestations and geographic distribution of loiasis overlap with those of other human filarial parasites, presenting challenges in the specific diagnosis of loiasis that may lead to delays in appropriate therapy. A recombinant antigen (Ll-SXP-1), preferentially recognized by serum samples from experimentally infected rhesus monkeys, was identified from a Loa loa L3 cDNA library. IgG4 antibody reactivity to purified Ll-SXP-1 was assessed by means of ELISA, using serum samples from patients with loiasis, lymphatic filariasis, onchocerciasis, mansonellosis, or other helminthiases and healthy control subjects. The assay was 56% sensitive and 98% specific for loiasis. Antibody reactivity was detectable before microfilaremia in experimentally infected rhesus monkeys and declined (but did not disappear) after diethylcarbamazine therapy in infected patients. IgG4 antibodies to recombinant Ll-SXP-1 are a highly specific marker of L. loa infection and may be useful for the diagnostic evaluation of persons with filariasis of unclear etiology.
Presented in part: 48th and 49th annual meetings of the American Society of Tropical Medicine and Hygiene, Washington, DC, November 1999 (abstract 210), and Houston, October 2000 (abstract 699), respectively.
Informed consent was obtained from all patients in accordance with US Department of Health and Human Services human experimentation guidelines.
Present affiliations: State University of New York Upstate Medical University, Syracuse, New York (A.V.); Harvard Medical School, Boston, Massachusetts (T.O.).
Reprints or correspondence: Dr. Amy Klion, National Institutes of Health, Laboratory of Parasitic Diseases, Bldg. 4, Rm. 126, Bethesda, MD 20892 .
Infection by the filarial parasite Loa loa can result in a wide range of clinical presentations, including asymptomatic microfilaremia, recurrent angioedema (Calabar swellings), subconjunctival migration of adult worms (eyeworm), and, rarely, severe complications, such as endomyocardial fibrosis and encephalopathy. Although definitive diagnosis of L. loa infection can be accomplished by identification of circulating microfilariae in a daytime blood sample or by excision of an adult worm, a majority (80%) of visitors to endemic areas and up to 20% of residents of endemic areas who acquire loiasis have no detectable parasites in the peripheral blood [1, 2]. The polymerase chain reaction is more sensitive than standard concentration methods, including Nucleopore filtration (Costar) and saponin lysis for parasitologic diagnosis of loiasis [3], but is impractical in field settings.
In the absence of parasitologic confirmation, diagnosis of loiasis hinges on a compatible clinical picture in the setting of eosinophilia and detectable antifilarial antibodies. This approach is complicated by the considerable serologic cross-reactivity and overlap in the clinical manifestations and geographic distributions of the filarial parasites that infect humans, including L. loa, Onchocerca volvulus, Wuchereria bancrofti, and Mansonella perstans. Despite the recent development of sensitive and specific diagnostic tests for onchocerciasis [4] and lymphatic filariasis [5, 6], coinfection with 2 filarial species is common in Africa. Thus, confirmation of filarial infection due to 1 of these species does not exclude concomitant loiasis.
L. loa is the only filarial parasite that infects humans for which there is curative therapy. Diethylcarbamazine (DEC), the drug of choice for the treatment of loiasis, is active against both the microfilarial and adult stages of L. loa and is curative in >95% of infected patients [7]. In contrast, patients with onchocerciasis are not cured by DEC treatment and may suffer severe side effects, including precipitation of blindness, after DEC administration. Conversely, ivermectin, the drug of choice for the treatment of onchocerciasis, is active only against the microfilarial stage of L. loa and is associated with severe adverse events, including death, in some patients with loiasis [8]. This has led to interruption of the mass distribution of ivermectin for control of onchocerciasis in some areas of central and western Africa. Thus, the lack of a specific marker of L. loa infection is problematic and may lead to delays in the initiation of appropriate therapy.
Until recently, the development of serologic assays for the diagnosis of loiasis was hampered by the lack of adequate parasite material. In this study, we used serum samples from rhesus monkeys experimentally infected with L. loa to screen an L. loa L3 cDNA library.
MATERIALS AND METHODS
Library screening. An L. loa L3 cDNA library constructed in ZAP II (Stratagene) was screened with pooled serum samples from 8 rhesus monkeys before and after L. loa experimental infection [9]. Filters were incubated sequentially with PBS0.1% Tween 20 (PBS-T) with 5% nonfat dry milk, Escherichia coliadsorbed patient serum (1 : 1000 in PBS-T with 5% nonfat dry milk), and horseradish-peroxidaseconjugated goat antihuman IgG (1 : 10,000 in PBS-T; Cappel). Positive plaques were detected by enhanced chemiluminescence (Amersham), according to the manufacturer's directions.
Dot blot analysis. Plaques preferentially recognized by the postinfection serum pool were purified and dotted onto a bacterial lawn. Protein expression was induced with an isopropyl--D-thiogalactopyranoside (IPTG)impregnated nitrocellulose membrane. The membranes were incubated sequentially and developed as described above.
Expression and purification of recombinant protein. Following rescue into the pBluescript phagemid by in vivo excision (Stratagene), plasmid inserts were sequenced from the 5 end, and database searches of the nucleotide and deduced protein sequences were performed by using BLAST at the National Center for Biotechnology Information [10]. A plasmid containing the presumed full-length coding region of Ll-spx-1, with the exception of 10 amino acids (aa) at the N-terminal end, was identified, purified, and digested with EcoRI and XhoI. The insert was gel purified and subcloned into the Topo TA vector (Invitrogen). The sequence of the plasmid insert was confirmed on both strands before digestion with BamHI and HindIII, gel purification, and directional cloning into pRSETB (Invitrogen) for overexpression. Recombinant plasmids were isolated, and inserts were sequenced to ensure the identity of the cloned inserts, proper orientation, and reading frame.
The recombinant plasmids were transformed into E. coli strain BL21(DE3), and protein expression was induced with 0.4 mM IPTG for 3 h. Histidine-tagged protein was purified over a nickel column (Ni-NTA agarose; Qiagen), according to the manufacturer's instructions, and was eluted with 100 mM imidazole in 20 mM Tris-HCl (pH 7.9). A recombinant protein of 21 kDa was detected on SDS-PAGE and confirmed by Western blot analysis by using E. coliadsorbed serum from a patient with loiasis, as well as antibody to the N-terminal histidine tag (data not shown).
ELISA. Preliminary studies were carried out to optimize the binding conditions and concentration of recombinant antigens for use in IgG and IgG4 antibody assays. An optimal concentration of 1 g/mL was selected for all subsequent studies. Microtiter plates (Immulon 2; Dynex) were coated overnight at 4°C with recombinant Ll-SXP-1 in PBS. Plates were blocked with PBS/5% bovine serum albumin (BSA)/0.5% Tween 20 for 2 h at 37°C before serum (diluted 1 : 75 in PBS/1% BSA/0.05% Tween 20) was added. After overnight incubation at 4°C, the plates were washed and incubated with mouse monoclonal antibody to human IgG4 (HP6024; CDC) and diluted 1:5000 in PBS/1% BSA/0.05% Tween 20 for 1 h at 37°C. Finally, the plates were washed and incubated with alkaline phosphataselabeled goat antimouse IgG Fc (Jackson Immunoresearch), diluted in the same buffer for 1 h at 37°C, and developed with alkaline phosphate substrate (Sigma-Aldrich) according to the manufacturer's instructions. All assays were done in duplicate, and the same positive reference serum was run on every plate in order to standardize the development time. A positive result was defined as a value >3-fold the standard deviation of the net optical density values for sera from patients without loiasis.
Patient serum. The endemic control serum and a subset of the filariasis serum were from previously described studies of loiasis [11, 12], lymphatic filariasis [13, 14], and onchocerciasis [15, 16]. The remainder of the filariasis serum, the serum from patients with other helminth infections, and the North American control serum, were from well-characterized patients seen at the National Institutes of Health. North American controls had no history of exposure to filarial or other nematode parasites.
RESULTS
Identification and preliminary characterization of Ll-sxp-1. An L. loa L3 cDNA library was screened using a serum pool from experimentally infected rhesus monkeys. Of 320,000 plaques screened, 14 were recognized by the rhesus monkey postinfection serum pool and not by the preinfection serum pool (data not shown). After plaque purification, antibody reactivity was reassessed by dot blot analysis by using human serum. All 14 clones were recognized by serum samples from 8 patients with loiasis but not by serum samples from the same 8 patients before infection or by serum samples from patients with lymphatic filariasis (n = 6), onchocerciasis (n = 6), or uninfected control subjects (n = 8; data not shown).
Preliminary sequence analysis revealed that 6 of the cDNA clones were identical and showed significant homology to Bm-sxp-1 from Brugia malayi (M98813 [17]). One of these, Ll-sxp-1 (AF174420), was sequenced in its entirety and was found to contain an insert of 629 bp. Based on comparison with the other known sxp-1like sequences (see [18] for alignment), Ll-sxp-1 is missing the coding region for 10 aa at the 5 end. A polyadenylation signal (ATTAAA) was found within the 179-bp 3 untranslated region, 11-bp upstream from the poly A tail. Ll-sxp-1 shows 32%52% identity at the aa level to other members of the filarial SXP/RAL-2 family of proteins (see ) with conservation of 32 of 34 conserved/invariant aa residues and both nematode-specific SXP motifs: (I/V)PPFLXGAPXXVV and (F/L)SP(E/A)A(R/K)XADAX(M/L)(S/T)AIA [18].
fig.ommitted
Figure 1. Comparison of predicted amino acid sequences of Ll-SXP-1 (accession no. AF174420), Ov17 (accession no. AI077025) [21], Wb-SXP-1 (accession no. AF098861) [18], and Bm-SXP-1 (accession no. M98813) [20] by multiple (Clustal) alignment. Shaded boxes indicate identical residues. Underlined areas are nematode-specific SXP motifs.
Antibody response to recombinant Ll-SXP-1. Histidine-tagged recombinant Ll-SXP-1 was purified over a nickel column for use in ELISAs, to determine the sensitivity and specificity of this assay in the diagnosis of L. loa infection. Preliminary experiments confirmed the presence of detectable Ll-SXP-1specific IgG in the serum of 20 of 24 patients with loiasis. However, Ll-SXP-1specific IgG also was detectable in a small number of patients with lymphatic filariasis (1/20) and onchocerciasis (4/20).
Antibody reactivity to Bm-SXP-1 was previously shown to be almost exclusively restricted to the IgG4 subclass [17] and a majority of patients with loiasis have elevated Loa-specific serum IgG4 levels [19]. We assessed the sensitivity and specificity of antiLl-SXP-1 IgG4 antibodies for the diagnosis of loiasis by ELISA by using serum samples from patients with loiasis (n = 48), onchocerciasis (n = 50), lymphatic filariasis (n = 49), or other filarial or helminth infections (n = 19) and from 25 healthy control subjects (16 from North America and 9 from Benin, western Africa; see and ).
fig.ommitted
Table 1. Patient population for serologic studies.
fig.ommitted
Figure 2. Detection of IgG4 antibodies to recombinant Ll-SXP-1 (Ll) in human serum (described in ) from healthy control subjects and patients infected with filarial and nonfilarial helminths. Each symbol represents the net optical density (OD) value (mean serum OD - mean diluent OD) for an individual patient. Shaded box represents a negative result, as defined by 3× the SD of values for serum samples from 124 subjects without loiasis. Ov, Onchocerca volvulus; Wb, Wuchereria bancrofti.
As shown in , 27 (56%) of the 48 patients with loiasis had detectable IgG4 antibodies to Ll-SXP-1. There was no association between antibody positivity and either the presence or the level of detectable blood microfilariae (data not shown). None of the control subjects or patients with lymphatic filariasis, mansonellosis, or nonfilarial helminth infections had detectable serum antiLl-SXP-1 IgG4 antibodies. AntiLl-SXP-1 IgG4 antibodies were detectable in the serum from 3 Guatemalan patients with onchocerciasis, albeit at low levels. None of the 3 patients had a history of travel to an area endemic for L. loa. Overall, the assay showed 98% specificity for the detection of L. loa infection with a positive predictive value of 90%.
The ability of the Ll-SXP-1 ELISA to detect prepatent infection was assessed by using serum samples from 4 rhesus monkeys experimentally infected with L. loa [9]. As shown in , IgG4 antibodies to Ll-SXP-1 were detectable in serum samples from 3 of the 4 monkeys at or near the onset of patency. The fourth monkey became seropositive 3 days after circulating microfilariae were detected in the peripheral blood.
fig.ommitted
Figure 3. IgG4 antibody reactivity against Ll-SXP-1 in rhesus monkeys experimentally infected with Loa loa. Each animal was inoculated with 75 infective third-stage (L3) L. loa larvae on day 0 and was monitored at least weekly for the appearance of microfilariae in blood [9]. Shaded boxes represent time points before the detection of circulating microfilariae (patency). All time points for an individual animal were run at the same time with a single microtiter plate. Each symbol represents the mean serum optical density (OD) value for a single time point in triplicate wells.
To determine whether the Ll-SXP-1 ELISA could distinguish between active and past infection, IgG4 antibody levels to Ll-SXP-1 were measured in the serum samples obtained from 8 patients with loiasis before treatment and were compared with serum samples from the same patients at the time of presumed cure. Cure was defined as resolution of all signs and symptoms after treatment with DEC without recurrence for 36 months, a normal peripheral blood eosinophil count (450/mm3), and decreasing IgE and antifilarial antibody levels. The median interval between samples was 7 years (range, 315 years). Although antiLl-SXP-1 IgG4 levels declined significantly (geometric mean, 0.246 vs. 0.078; P < .01), antibodies remained detectable in 4 of 8 patients at the time of presumed cure .
fig.ommitted
Figure 4. Decline in IgG4 antibody reactivity to Ll-SXP-1 in patients with loiasis before treatment and at the time of presumed cure. Each symbol represents the mean serum optical density (OD) value for a single time point in triplicate wells. Lines connect the pretreatment and follow-up values for a single patient (median time interval between serum samples, 7 years; range, 315 years).
DISCUSSION
The development of rapid diagnostic assays for onchocerciasis [4] and lymphatic filariasis [5, 6] has simplified the care of individual patients with these filarial infections and allowed for accurate and cost-effective geographic mapping for the purposes of mass chemotherapy in endemic areas. Nevertheless, the coendemicity of loiasis with other human filarial infections remains an issue because of species-specific differences in responses to available antifilarial therapies. This is of particular concern in the setting of onchocerciasis control programs, where several deaths have occurred among Loa-infected patients who received ivermectin [8]. With the expansion of ivermectin distribution programs in Africa to areas endemic for lymphatic filariasis, accurate mapping of areas coendemic for loiasis will be even more important.
Ll-SXP-1 belongs to the Sxp1/RAL family of nematode proteins. Members of this protein family were previously identified by differential screening of cDNA libraries from several different life-cycle stages of B. malayi [17, 20], W. bancrofti [18], and O. volvulus [21]. Despite significant homology among these filarial SXP-1 orthologues at the amino acid level, the human IgG antibody responses to these proteins demonstrate a high degree of species specificity that is enhanced when antigen-specific IgG4 responses are measured [18]. Similar to the results obtained with other SXP-1 proteins, the IgG4 antibody response to Ll-SXP-1 is highly specific (>99%) for L. loa infection. The relative lack of sensitivity (56%) of this response in detecting patients with loiasis may reflect HLA restriction of the IgG4 response to Ll-SXP-1, as reported with other recombinant antigens, or the constraints of the assay.
In lymphatic filariasis, species-specific IgG4 antibody levels are elevated in active infection, as defined by the presence of detectable blood microfilariae, and are correlated with the level of microfilariae detectable in the peripheral blood (microfilaremia) [22, 23]. Although studies that used crude L. loa adult antigen demonstrated that Loa-specific IgG4 is a more specific marker of L. loa infection than Loa-specific total IgG [24], Loa-specific IgG4 was elevated in both microfilaremic and amicrofilaremic persons, and levels showed no correlation with microfilaremia [25]. IgG4 antibody reactivity to Ll-SXP-1 showed a similar pattern to that observed with crude antigen. This and the fact that antibodies to Ll-SXP-1 were detected before the detection of microfilaremia in 3 of 4 rhesus monkeys suggest that the immune response Ll-SXP-1 may not be directed exclusively at the microfilarial stage of the parasite. Protein expression of Bm-SXP-1 has been confirmed in both the adult and microfilarial stages of the parasite and in their excretory/secretory products, which is consistent with this hypothesis [17].
One problem with antibody detection in chronic infection is the inability to distinguish active from past infection, because of the persistence of detectable antibodies over long time periods. In patients with W. bancrofti infection, IgG antibodies to Bm-sxp-1 declined slowly over time after treatment with DEC [17]. A similar decline was seen in the IgG4 antibody levels to Ll-sxp-1 in DEC-treated patients, although antibody levels remained elevated in a majority of patients, despite presumed cure of infection (the absence of signs or symptoms consistent with infection and no history of travel to an area endemic for loiasis in >10 years). This is of concern both in epidemiologic screening and in the treatment of individual patients with loiasis, because there are no clinical or laboratory predictors of treatment failure and recurrences may occur after long asymptomatic intervals. As with lymphatic filariasis, the development of assays to detect circulating parasite antigens would provide a much-needed marker of active infection. The identification of recombinant L. loa antigens that induce a species-specific antibody response, including Ll-SXP-1, is the first step in the development of reagents necessary for such assays.
In summary, antiLl-SXP-1 IgG4 is a highly specific marker of L. loa infection. Despite relatively low sensitivity (56%), this assay should be useful in the identification of areas endemic for loiasis. Because of the high predictive value of a positive result, the Ll-SXP-1 ELISA also should allow for confirmation of L. loa infection in select clinical settings. Ultimately, the development of an assay to detect circulating Ll-SXP-1 may increase the sensitivity of detection of L. loainfected persons and provide a means of monitoring the success of chemotherapy.
Acknowledgments
We thank Joseph Kubofcik, Joseph Tropea, and Nithyakalyani Raghavan for helping with the expression and purification of the recombinant protein and Mark Eberhard for providing rhesus monkey serum samples.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.