A comparative study of two different assay kits for the detection of secreted alkaline phosphatase in HPV antibody neutralization assays

Abstract

To assess immunogenicity and development of antibodies in the context of vaccination, it is critical to quantify titers of neutralizing antibodies. We have been employing the 293TT cell-based neutralization assay system to quantify anti-HPV neutralizing antibodies. In this system, human papillomavirus (HPV) pseudovirion (PsV) particles encapsidating secreted alkaline phosphatase (SEAP) gene are used to measure infection of 293TT cells in 72-hr cell-culture supernatants. SEAP has traditionally been measured by Great EscAPe™ SEAP Chemiluminescence Kit 2.0 (GE). To reduce the cost, and to potentially increase efficiency, we sought a cheaper kit with better detection capability. Performance characteristics of the newer chemiluminescence kit, ZiVa® Ultra SEAP Plus Assay (Ziva) and GE were compared using the 293TT system. Dose titration of HPV PsV 16 or 18 showed that signal-to-noise ratios at 48 and 72 hr post-infection were higher for ZiVa at nearly all doses. ZiVa was superior to GE as it was able to detect SEAP at 48 hr, as well as when lower numbers of 293TT cells were used. The ability of ZiVa to quantitate HPV-16 and -18 neutralizing antibody titers was tested using sera from Cervarix® immunized individuals. Spearman rank correlational analyses showed excellent correlations between the titers obtained with ZiVa and GE for anti-HPV16 (r = 0.9822, p < 0.0001) and anti-HPV18 (r = 0.9832, p < 0.0001) antibodies. We concluded that ZiVa is superior to GE in detecting SEAP, and the antibody titers in sera of vaccinated individuals were similar to those obtained with GE. Thus, Ziva is a suitable alternative to GE.

Keywords:

• Cervarix®
• Gardasil®
• human papillomavirus
• SEAP
• vaccine
• ZiVa

Abbreviations:

• HPV, human papillomavirus
• VLP, virus-like-particle
• PsV, pseudovirion
• ZiVa, ZiVa Ultra SEAP Plus Assay
• GE, Great EscAPe™ SEAP Chemiluminescence Kit 2.0
• ELISA, enzyme-linked immunosorbent assay
• GFP, green fluorescent protein
• L1 and L2, late 1 and late 2 proteins
• RLU, relative-light-unit

Introduction

The most common sexually transmitted viral infections are caused by human papillomaviruses (HPV).^1,2 HPV are also the major etiological agents responsible for development of benign genital warts (types 6 and 11), precancerous and cancerous lesions of the cervix and for a portion of anogential and oral malignancies (types 16 and 18).^3-7 HPV types 16 and 18 are responsible for approximately 70% of cervical cancers.⁸ In 2006 and in 2009, the United States Food and Drug Administration (FDA) approved 2 different prophylactic HPV vaccines: one of them is for females and males,⁹ and the other for females only.¹⁰ These 2 HPV vaccines, Gardasil® and Cervarix®, are comprised of late 1 (L1) major capsid protein-based virus-like-particles (VLPs) of HPV6, 11, 16 and 18 (Gardasil®) or HPV16 and 18 (Cervarix®).¹¹ These vaccines have demonstrated excellent efficacy in clinical trials. In response to these L1 VLP vaccines, virus-neutralizing L1-specific antibodies are generated, whose levels and neutralization activity can be quantitated by enzyme-linked immunosorbent assay (ELISA) and in vitro neutralization assay, respectively.

In addition to these licensed vaccines, second-generation HPV vaccines are in development. For example, the nonavalent L1 VLP vaccine that contains VLPs for HPV types 6, 11, 16, 18, 31, 33, 45, 52, and 58 has been developed to widen the range of protection against other carcinogenic HPV types,¹¹ and it is under review for licensure by the FDA. Another second-generation vaccine that is based on the minor capsid protein L2 has also been in the development because of L2's potential to be broadly reactive;¹² thereby providing a wider range of protection against different HPV types.^11,13 A recombinant fusion protein composed of amino acid sequence positions 11 to 88 of L2 proteins has shown promising preclinical results.^14,15 However, L2-based vaccines have yet to be evaluated in clinical trials.

A measure of the success of a given vaccine is dependent on its ability to provide protection from the targeted infectious agent and the disease that it causes. Although no correlates of protection have formally been identified for the HPV vaccines, neutralizing antibodies are believed to play a key role in providing protection against HPV infection. Therefore, quantitation of titers of biologically active antigen-specific antibodies becomes critical in understanding the development of humoral immunity in response to a vaccine. As such, cell-based in vitro neutralization assay is an ideal system to measure the levels of biologically active antibodies in clinical samples. Although different neutralization assays have been reported for HPV,^16-25 our laboratory has adopted, optimized, and validated^26-29 a pseudovirion (PsV)-based HPV neutralization assay that was originally reported by Pastrana et al.³⁰ In this assay system, PsV particles that are composed of L1 and L2 capsid proteins from a particular HPV type are used to infect 293TT cells for 72 hr. Because a PsV particle encapsidates a reporter gene during its production,^31,32 magnitude of PsV entry and infection can be measured through the detection of the reporter gene product. Using this 293TT cell-based system and secreted alkaline phosphatase (SEAP) as a reporter, we have measured titers of biologically active anti-L1 specific antibodies in HPV L1VLP vaccine recipients as well as in those who were naturally infected with HPVs.^27,28,30

As the number of samples from clinical trials can be large, conducting the cell-based neutralization assay and SEAP assay can be labor-intensive and expensive. This 293TT cell-based neutralization assay takes 72 hr before collecting and storing cell-culture supernatants for the subsequent SEAP assay. Furthermore, production of and the use of PsVs in the neutralization assay can also be costly when testing hundreds to thousands of clinical samples. Hence, it is desirable to use the lowest amount possible of virus particles that would still lead to a reasonable induction of signal over background. For these reasons, we sought a cheaper SEAP detection kit with a better detectability to reduce the expenditures of resources and to increase efficiency. However, we wanted to ultimately examine whether such kit can still be used in measuring HPV neutralizing anti-L1 antibodies in clinical serum samples.

We and others have previously used Great EscAPe™ SEAP Chemiluminescence Kit 2.0 from Clontech (GE) for the detection of SEAP,^28-30 while a different group of investigators has used the SEAP Reporter Gene Assay for Chemiluminescence from Roche.³³ In addition to these 2 well-known brand name assay kits, we found 2 additional products after performing on-line searches. One of them, ZiVa® Ultra-Sensitive SEAP Detection Kit (ZiVa) from Jaden Bioscience, Inc., appeared to fit our need of a kit that is less expensive and highly sensitive. In this study, we tested whether ZiVa can be an alternative product to GE. We conducted a side-by-side comparison of ZiVa and GE using the 293TT cell-based assay system. This comparison included evaluation of kits’ ability to detect SEAP through performing infectivity assays using different concentrations of PsV particles of HPV16 and 18, and by using different numbers of target 293TT cells. In addition, SEAP activity was examined at 48 and 72 hr post-infection to evaluate whether the duration of 293TT cell-based assay could be shortened. Most important, we examined the ability of ZiVa to quantitate HPV neutralizing titers in sera from a subset of Cervarix® immunized individuals who received 1, 2, or 3 doses of the vaccine as part of a Uganda-government sponsored vaccine demonstration program,³⁴ and compared how well these titers correlated with those obtained with the GE kit.

Results

Comparison of different SEAP assay kits

To find an alternative to Clontech's Great EscAPe^™ SEAP Kit 2.0 (hereon, referred to as GE), we searched for SEAP chemiluminescence kits from different vendors and compared their characteristics (Table 1). In addition to GE, we found kits from Roche, Life Technologies, and Jaden Bioscience, Inc. All four vendors offered product size of approximately 1,000 tests, except for Roche, which had 500 tests as its biggest size. Based on the available information from the manufacturers, we found that the majority of the characteristics were similar between the kits (Table 1). However, the product from Jaden Bioscience had other unique, attractive features. These included lower cost and better detectability of SEAP (Table 1). These characteristics led us to compare the ability of ZiVa® Ultra SEAP Plus Detection Kit (hereon, referred to as ZiVa) and GE to detect SEAP in supernatants from cell-based bioassays.

Table 1. Comparison of different chemiluminescence-based SEAP assay kits

	VENDOR
Characteristics^a	Clontech	Roche	Life Technologies	Jaden Bioscience
Catalog no.	631738	11779842001	N10578	CM025
Kit size	1000 tests	500 tests	10 x 96 well plates	1000 tests
96-well plate format	Yes	Yes	Yes	Yes
Storage temperature	−20°C	4°C	4°C	4°C
List price (US $)	$838.00	$513.00	$1,124.00	$389.99
Sensitivity	100 fg	10 fg	NA^b	0.005 fg
Linear dynamic range^c	10^4	10^4–10^5	10^5	10^6 –10^7
Sample volume used (μL)	25	50^d	25	1–5
Approximate assay time (min)	60	60	30	60

^aThe information is based on those indicated in the production sheets or certificate of analysis, or on the website of each company. ^bNA: Not available. ^cRange of detection will vary depending on the instrument used to measure the activity. ^dFifty microliters of samples from 4-fold dilution are used.Website for each company: Clontech (http://www.clontech.com/US/Products/Fluorescent_Proteins_and_Reporters/Chemiluminescent_Reporters/Secreted_Alkaline_Phosphatase); Roche (http://www.roche-applied-science.com/shop/products/seap-reporter-gene-assay-chemiluminescent#tab-0); Life Technologies (http://www.lifetechnologies.com/us/en/home/life-science/protein-expression-and-analysis/enzymes-and-protein-activity-assays/epaassays-misc/novabrightchemiluminescent-reporter-gene-assays/novabright-chemiluminescent-seap-reporter-gene-assays.html); and Jaden Bioscience (http://www.jadenbio.com/assays/protein/ziva-seap.html)

ZiVa is superior to GE in its ability to detect SEAP in cell-culture supernatants

To compare the ability of ZiVa and GE to detect SEAP levels, 293TT cells were infected with SEAP gene encapsidated HPV pseudovirion (PsV)16 or PsV18 particles that were 2-fold serially diluted (range: 1,000 – 1,024,000). To express the magnitude of infections, signal-to-noise ratio (fold-increase) was calculated by dividing the relative-light-unit (RLU) of an infected sample by the mean RLU of the no-virus controls, in which 293TT cells were incubated without any viral particles. The mean fold-increase was calculated using individual values of the duplicate wells, and the results of 3 independent experiments for both PsV16 and PsV18 are shown in Fig. 1.

Figure 1. ZiVa can better detect SEAP than GE. Pseudovirion particles of HPV types 16 and 18 were serially diluted, and used to infect 0.03 × 10⁶ 293TT cells. Two-fold serial dilutions were performed starting from 1:1000 to 1:1024000. Supernatants were harvested at 48 and 72 hr post-infection. Each PsV dilution was setup in duplicate wells. Mean ± standard deviation (SD) of 3 independent experiments are shown. (A) Infectivity levels (fold-increase) by PsV16 particles after 48 or 72 hr of incubation are shown. (B) Infectivity levels by PsV18 particles after 48 or 72 hr of incubation are shown. In both (A and B), the same supernatant samples were tested using ZiVa or GE SEAP detection kit. The plots show fold-increase in log₁₀ scale as a function of viral concentration (log₁₀). Some of the error bars are too small to be seen in the graphs.

For both PsV16- (Fig. 1A) and PsV18-infected (Fig. 1B) samples, ZiVa out performed GE at both 48- and 72-hr time points, despite using only 5 μL of the supernatants vs 25 μL of GE. With ZiVa, higher fold-increases were achieved for all dilutions tested for PsV16 at both time points with the mean ZiVa/GE ratios that ranged from 16–24 at 48 hr, and 10–16 at 72 hr (Table 2). For PsV18, the mean ratios ranged from 1–26 and 12–16 at 48 and 72 hr, respectively (Table 2). The last 2 PsV18 viral dilution factors (512,000 and 1,024,000) showed a minimal to no difference at the 48-hr time point for PsV18. A further indication that ZiVa can better detect SEAP than GE was that when we arbitrarily chose to examine the amount of viral particles required to induce 100-fold increase in signal, we found that ZiVa required at least 10-fold less particles than those required for GE (Fig. 1, dotted lines). At 72 hr post-infection, the values of fold-increase for PsV16 remained above 100-fold for ZiVa even at the lowest dilution of 1:1,024,000 (Fig 1A, right panel). Furthermore, mean RLUs of the no-virus controls (n = 8 wells per experiment) from the 3 experiments at 48 hr were 57, 67, and 73 (mean = 66), and were 118, 143, and 168 (mean = 143) at 72 hr for ZiVa. For GE, they were 170, 190, and 228 (mean = 196) at 48 hr, and 241, 278, and 247 (mean = 255) at 72 hr. The values were higher for GE by factors of 3.0 (196/66) and 1.8 (255/143) for 48- and 72-hr time points, respectively. Therefore, the higher background RLU for GE does not explain better fold-increases obtained by using ZiVa.

Table 2. Ratios of ZiVa to GE in detecting SEAP

	ZiVa/GE ratio (48 hr)				ZiVa/GE ratio (72 hr)
PsV16 dilution	Exp 1	Exp 2	Exp 3	Mean(48 hr)	Exp 1	Exp 2	Exp 3	Mean (72 hr)
1,000	21	22	23	22	12	11	8	10
2,000	22	23	24	23	12	11	9	11
4,000	22	24	26	24	13	12	9	12
8,000	23	23	25	24	15	14	10	13
16,000	23	25	24	24	15	14	11	14
32,000	23	24	26	24	17	16	12	15
64,000	23	23	25	23	18	16	12	16
128,000	22	25	25	24	19	17	12	16
256,000	21	22	24	22	18	16	13	15
512,000	18	24	23	22	18	17	12	16
1,024,000	12	19	18	16	17	15	12	15
PsV18 dilution	Exp 1	Exp 2	Exp 3	Mean (48 hr)	Exp 1	Exp 2	Exp 3	Mean (72 hr)
1,000	23	24	27	24	15	14	10	13
2,000	23	24	26	25	16	15	11	14
4,000	24	25	28	26	17	17	12	15
8,000	24	24	28	26	18	17	12	16
16,000	24	22	27	24	18	18	12	16
32,000	23	22	24	23	19	18	12	16
64,000	12	13	24	16	19	18	12	16
128,000	12	11	13	12	19	18	12	16
256,000	9	7	7	8	17	17	12	15
512,000	1	4	1	2	12	14	11	12
1,024,000	2	1	2	1	15	15	8	12

Values in the ZiVa/GE ratios were calculated by dividing the values of mean fold-increase obtained using ZiVa by the values obtained with GE for the corresponding PsV dilution factors.

To examine whether there was an added benefit to culturing 72 hr when compared to 48 hr, we divided the values of mean fold-increase of the 72-hr time point by those of the 48-hr time point within each assay for both PsV16 and PsV18 (Tables S1 and 2). For ZiVa, the overall ratios of 72-hr to 48-hr time points ranged from 1 to 19 over 3 independent experiments for PsV16, while the range was 3 to 18 for GE (Table S1). For PsV18, they were 6 to 106 for ZiVa, and 3 to 30 for GE (Table S2). As expected, with higher dilutions, the ratios were higher for both PsV types. However, because the last 5 dilutions were border line limiting factors for PsV18, the ratios were skewed toward higher values. Overall, there was an added benefit to culturing for 72 hr for both ZiVa and GE.

Lower numbers of 293TT cells can be used to detect SEAP with ZiVa

We further tested the ability of the 2 kits to detect SEAP by using different numbers of 293TT cells per well (range: 0.002 × 10⁶ to 0.3 × 10⁶). For this study, we used 2 different viral concentrations for PsV16 (1:100,000 and 1:600,000), and for PsV18 (1:10,000 and 1:60,000) to infect different numbers of 293TT cells. These viral dilutions factors were expected to provide approximately 100,000 to 300,000 RLUs using either ZiVa or GE based on the results from the infectivity assay (Fig. 1). The lower dilution factors were to provide the expected RLU values when using GE, and the higher dilutions factors were for ZiVa. Cell-culture supernatants were collected as in Figure 1, and the results from 3 independent experiments are shown in Figures 2 and 3.

Figure 2. ZiVa can detect SEAP even using limiting numbers of 293TT cells. Different numbers of 293TT cells were plated at the start of the infections by PsV16 or PsV18. The number of cells ranged from 0.002 - 0.3 × 10⁶ cells per well, and 2-fold serial dilutions were performed. Each cell number was tested in triplicates. A fixed dilution factor of 1:100000 for PsV16 (A) and 1:10000 for PsV18 (B) were used to infect 293TT cells. At 48 or 72 hr post-infection, the supernatants were harvested. Cell numbers were plotted in log₁₀ scale. Mean ± SD of 3 independent experiments are shown. Some of the error bars are too small to be seen in the graphs.

Figure 3. ZiVa can detect SEAP at lower cell numbers using lower viral dilution concentrations. The experiments were conducted as described in Fig. 2, except higher viral dilution factors were used: 1:600000 for PsV16 and 1:60000 for PsV18. Mean ± SD of 3 independent experiments are shown. Some of the error bars are too small to be seen in the graphs.

In general, the magnitude of signal increased until the cell number reached 0.038 × 10⁶ cells per well, and then, it began to decline (Figs. 2 and 3). The maximal fold-increases were detected at 0.038 × 10⁶ cells per well for both HPV16 (Figs. 2A and 3A) and HPV18 (Figs. 2B and 3B), regardless of the viral concentrations. However, the ability of GE to detect SEAP was inferior to ZiVa at both 48- and 72-hr time points; for the majority of cell numbers tested, the magnitude of fold-increase was much less than those of ZiVa.

Overall, the data showed that ZiVa is capable of detecting the amount of SEAP that is limiting for GE in cell-culture supernatants, and that ZiVa can provide higher level of signal than GE. As a result, lower amount of PsV can be used and still obtain a high signal-to-noise ratio. Furthermore, although we observed an optimal signal induction at 0.038 × 10⁶ cells per well, the data showed that ZiVa can detect substantial amount of the enzyme even when using lower numbers of cells.

HPV-specific antibody titers measured by ZiVa correlate with those of GE

HPV immunogenicity studies often involve testing of large numbers of clinical serum samples from HPV vaccinated individuals to determine neutralizing antibody titers. In our previous studies, we used GE to conduct SEAP assay to measure HPV neutralizing antibody titers.^26,29,35 To determine if ZiVa could be an alternative to GE in clinical studies, we evaluated whether the neutralization titers measured by ZiVa correlate to those measured by GE. For this, we used culture supernatants from 293TT cell-based neutralization assays that were performed for the Uganda Immunogenicity Study.³⁶ Anti-HPV16 and -HPV18 antibody titers were determined using randomly selected sera from those who received 1 (n = 25), 2 (n = 24), or 3 doses (n = 23) of the vaccine.

The results showed excellent overall correlations between the values obtained by GE and ZiVa for both anti-HPV16 and -HPV18 antibody titers (p < 0.0001 for both), with r = 0.9822 for anti-HPV16, and r = 0.9832 for anti-HPV18 (Fig.4). We further tested for the correlations within the 3 different dosage groups. Regardless of the number of doses received, we found significant correlations (p < 0.0001 for all) between the values of titers measured by ZiVa and GE for anti-HPV16 (r = 0.9592, 0.9730, and 0.9427 for 1, 2, and 3 doses, respectively), and for anti-HPV18 antibodies (r = 0.9862, 0.9760, and 0.9753 for 1, 2, and 3 doses, respectively). The anti-HPV16 and -HPV18 antibody titers in these serum samples covered wide ranges; they were from 1:27 to 1:109,285 for anti-HPV16 for ZiVa, and from 1:16 to 1:17,828 for GE. For anti-HPV18, they were 1:9 to 1:8,745 for ZiVa, and 1:8 to 1:7667 for GE. These data showed that ZiVa can be used to quantitate the titers of samples from vaccinated individuals, demonstrating the feasibility of its use in clinical studies. Moreover, when we compared the values of 50% neutralization titers measured by GE or ZiVa, we found that the titer values were very similar, except for sample ID 64 for anti-HPV16 (Table 3). The data in Table 3 provide further support that ZiVa can be used to measure the antibody titers in clinical serum samples that are comparable to the ones obtained with GE.

Table 3. Comparison of 50% neutralization titer values measured by GE or ZiVa

	Anti-HPV16			Anti-HPV18
Sample ID	GE	ZiVa	ZiVa/ GE^a	GE	ZiVa	ZiVa/ GE
1	129	175	1.4	373	432	1.2
2	207	162	0.8	112	169	1.5
3	343	294	0.9	302	495	1.6
4	124	126	1.0	180	300	1.7
5	6999	5927	0.8	1112	1127	1.0
6	562	507	0.9	1301	2509	1.9
7	404	444	1.1	116	113	1.0
8	144	563	3.9	80	95	1.2
9	47	29	0.6	98	129	1.3
10	180	331	1.8	127	148	1.2
11	354	391	1.1	159	148	0.9
12	45	81	1.8	126	187	1.5
13	3602	2474	0.7	2279	2705	1.2
14	2491	2428	1.0	3440	4122	1.2
15	737	861	1.2	192	314	1.6
16	62	57	0.9	11	11	1.0
17	1357	1292	1.0	588	563	1.0
18	278	339	1.2	61	80	1.3
19	853	926	1.1	349	469	1.3
20	1139	1025	0.9	805	755	0.9
21	633	664	1.0	442	498	1.1
22	5159	4304	0.8	1261	1202	1.0
23	16	27	1.7	8	9	1.2
24	312	283	0.9	297	370	1.2
25	4063	4301	1.1	3463	3207	0.9
26	4098	3657	0.9	1900	1903	1.0
27	239	253	1.1	344	295	0.9
28	3378	4287	1.3	1846	1638	0.9
29	3795	3690	1.0	3704	3407	0.9
30	6256	5214	0.8	2784	3520	1.3
31	6060	6596	1.1	3213	2852	0.9
32	6725	4783	0.7	1724	1716	1.0
33	127	175	1.4	53	77	1.5
34	695	841	1.2	269	344	1.3
35	2339	2835	1.2	604	786	1.3
36	7341	6794	0.9	2292	2504	1.1
37	5394	4429	0.8	3186	2806	0.9
38	4409	4069	0.9	6670	6142	0.9
39	4791	6364	1.3	6013	6455	1.1
40	1898	1834	1.0	1282	1430	1.1
41	8565	9831	1.1	5761	6526	1.1
42	245	372	1.5	259	289	1.1
43	585	554	0.9	399	418	1.0
44	3930	3445	0.9	3484	2479	0.7
45	2727	2966	1.1	859	1052	1.2
46	4477	4767	1.1	879	774	0.9
47	9393	10044	1.1	1767	2038	1.2
48	305	222	0.7	649	821	1.3
49	1843	1958	1.1	670	711	1.1
50	8215	7633	0.9	2022	1996	1.0
51	5760	5838	1.0	894	821	0.9
52	1187	1035	0.9	1351	1292	1.0
53	5740	4756	0.8	369	378	1.0
54	492	534	1.1	45	46	1.0
55	8855	8281	0.9	7667	8745	1.1
56	1065	1043	1.0	458	561	1.2
57	5213	5355	1.0	292	281	1.0
58	1217	942	0.8	142	423	3.0
59	2624	1980	0.8	1825	2381	1.3
60	1675	1383	0.8	550	773	1.4
61	3106	2451	0.8	1977	2443	1.2
62	17828	24856	1.4	4312	5285	1.2
63	360	528	1.5	99	92	0.9
64	5054	109285	21.6	2209	3339	1.5
65	17124	14622	0.9	1910	1841	1.0
66	595	542	0.9	526	562	1.1
67	2361	2599	1.1	1905	1897	1.0
68	6893	7705	1.1	2264	2480	1.1
69	3321	2901	0.9	400	340	0.9
70	3068	3275	1.1	491	770	1.6
71	4170	3494	0.8	387	374	1.0
72	6826	6637	1.0	1376	1655	1.2

Fifty percent neutralization titers measured by ZiVa or GE from the corresponding samples are shown.

^aThe values of 50% neutralization titers measured by ZiVa were divided by those measured by GE.

Figure 4. HPV neutralizing antibody titers measured by ZiVa correlate with those of GE. Cell-culture supernatants from 293TT cell-based neutralization assay from the Uganda Immunogenicity Study were used to quantitate 50% neutralization titers in sera of Cervarix®-immunized individuals using ZiVa or GE. The supernatant samples collected after 72 hr of incubation were used to measure the SEAP activity. Spearman rank-correlational analyses were performed between the titers obtained by ZiVa or GE for both anti-HPV16 and -HPV18 antibodies.

Together, the data showed that HPV neutralizing titers in serum samples from Cervarix® immunized individuals determined by ZiVa significantly correlated to the titers obtained by GE, regardless of the number of doses received. In addition, amounts of PsV particles required to obtain similar levels of fold-induction were at least 10-fold less for ZiVa compared to GE. Therefore, the ability of ZiVa to detect SEAP is superior to that of GE, and that ZiVa can serve as an alternative SEAP detection kit with the potential to substantially lower the cost of immunogenicity studies of HPV vaccines.

Discussion

In this study, we compared the ability of Ziva® Ultra SEAP Plus Assay and Great EscAPe™ SEAP Chemiluminescence Kit 2.0 to detect SEAP. Because the product information for ZiVa indicated that it was more sensitive (at polymerase chain reaction [PCR]-level, per Jaden BioScience, Inc.) and cheaper than GE, we wanted to compare the ability of the 2 kits to detect SEAP in cell-culture supernatants. However, our major aim was to examine whether ZiVa could measure HPV neutralizing antibody titers in sera of vaccinated individuals at comparable levels to those measured by GE. We found that ZiVa was able to detect SEAP better than GE, and that it quantitated the titers of anti-HPV antibodies that highly correlated to those measured by GE.

Although we used SEAP as a reporter gene, green fluorescent protein (GFP),^30,37 Gaussia luciferase,³³ or firefly luciferase³⁸ has been used in the 293TT cell-based system as an alternative reporter gene. Our laboratory has chosen SEAP as the reporter gene of choice because of its high sensitivity as well as its potential application in a high-throughput system. GFP is useful in quantitating the number of cells that were infected by PsV particles, while the luciferase system can also be used to measure in vivo infection in animal models. However, GFP and firefly luciferase require collection or lysis of cells for subsequent measurement by a flow cytometer or by a luminometer with an injector, respectively. In contrast, the activity of SEAP can directly be measured using only a few microliters of cell-culture supernatants, and does not require an injector. Although Gaussia luciferase can also be detected using cell-culture supernatants, it still requires a luminometer with an injector. Furthermore, an entire 96-well plate can be read in seconds to minutes for SEAP, and once a maximal level of signal is reached, it is stably maintained for hours. In contrast, signals generated by luciferase starts to decay within seconds to minutes, and collection of cells by a flow cytometer can take several minutes to hours even if the samples were measured in a plate-format.

Our data showed that less amounts of PsV particles can be used with ZiVa, which could potentially enable the use of the same lot of viral particles for multiple studies. Our data from infection assays demonstrated that ZiVa can detect SEAP even at the PsV doses that were limiting for GE. However, when we compared the ratios of signals by dividing those from the 72-hr time point by those from the 48-hr time point, we found similar magnitude of increases for both kits. We concluded that there was an added benefit to culturing 72 hr. However, because of ZiVa's superior detectability, the assay could potentially be terminated after 48 hr of culture. In addition, although we observed that the optimal fold-increase was obtained with 0.038 × 10⁶ cells per well, we did not observe much of a difference when half the number of cells (0.019 × 10⁶ cells per well) was used, indicating that the cell number could be lowered, which could reduce the cost associated with cell culture.

Although Sehr et al.³³ have reported that Gaussia luciferase in their high-throughput system was more sensitive than SEAP, they used the reporter assay kit from Roche. In addition, ZiVa is supposed to provide a PCR-level of sensitivity. Nevertheless, a future, formal comparison of SEAP assay detection kits from Roche and Jaden BioScience Inc. may be needed, along with Gaussia luciferase in the context of vaccination and/or natural infection.

We found significant correlations between the titers obtained by both kits. The correlations were equally high regardless of whether the participants had received 1, 2, or 3 doses. Although we did not test sera from naturally infected individuals, whose titers are expected to be lower than those of vaccinated individuals, the samples tested in this study covered a wide range of titers, including some that were near the cut-off titers (1:10). Hence, we expect that ZiVa will also be applicable to epidemiological studies of HPV natural infection.

Correlational analyses showed that there was a high agreement (r > 0.98 for both HPV16 and 18) between titers of antibodies spanning a wide range of levels induced by 1, 2, or 3 doses; indicating that the performance of the assay is independent of the levels of HPV neutralizing antibodies. Although one of the samples (ID64) showed a higher titer value for ZiVa, we believe this is an outlier. As the titers for anti-HPV18 were similar in this sample, we will need to increase the study samples size to determine the frequency of such event. Of note, because ZiVa could allow the use of lower concentrations of PsV particles, it is possible that the usage of lower numbers of PsV particles would change the values of 50% neutralization titers. However, this should not change the overall results and interpretation of the data, such that samples with high or low titers will be expected to be ranked as such. Nevertheless, formal future studies may be warranted using a lower PsV concentration with vaccinated and/or naturally infected serum samples.

In conclusion, we believe that ZiVa can be an alternative to GE, because of its superior detectability of SEAP and its lower price. In addition, our data showed that the length of the experiments could potentially be shortened, and a lower number of cells could be used. Both of these parameters need to be further evaluated. They may further increase the efficiency of the experiments, and add to the system that is already high-throughput in the generation of data.

Materials and Methods

Cell culture and culture media

293TT cells have previously been described.³² The cells were maintained in Dulbecco's Modified Eagle Medium (DMEM) (11965; Gibco/Life Technologies) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (SH30070/03; Hyclone), 1% MEM non-essential amino acids (H11140–050) and 1% Glutamax (35050–061) (both from Gibco/Life Technologies). To maintain the expression of large-T antigen, the cells were cultured in the presence of 400 μg/mL hygromycin B (10687–010; Gibco/Life Technologies). For 293TT cell infectivity assay and 293TT cell-based neutralization assay, phenol red-free DMEM (21063–029; Gibco/Life Technologies) supplemented with above reagents were used along with 1% antibiotic/antimycotic (15240–062) and 1% HEPES (15630–080) (both from Gibco/Life Technologies); however, hygromycin B was not included in the media.

Production of PsV particles with SEAP reporter gene

The production of PsV particles has previously been reported.³² Briefly, for the production of PsV particles with encapsidated SEAP gene, 293TT cells were transfected with a pShell plasmid that contains codon-modified L1 and L2 genes³² of HPV16 or HPV18 with pYSEAP plasmid ³⁰ using Lipofectamine 2000 (11668–019; Invitrogen/Life Technologies). The cells were cultured for 2 d in a 37°C CO₂ incubator after the transfection. To collect the PsV particles produced inside 293TT cells, the cells were harvested, lysed in Lysis Buffer (PBS with 10 mM MgCl₂ [14040–141; Gibco/Life Technologies], 0.5% Brij58 [P-5884; Sigma], and RNase A/T1 cocktail [AM2286; Ambion]) without Benzonase. Released PsV particles in Lysis Buffer were matured³⁹ by incubating them in a 37°C water bath for 24 hr. The matured PsV particles were purified through Optiprep (D1556; Sigma) gradient (27%, 33%, and 39%) by ultracentrifugation at 50000rpm (Beckman Coulter Optima^™ L-80 XP) for 3.5 hr. Purified PsV particles were collected, aliquoted into siliconized tubes (4203SLS, Bio Plas, Inc.), and stored at −80°C until use.

293TT infectivity assay

To perform infection of 293TT cells with PsV16 or PsV18 particles encapsidating the SEAP gene, 0.03 × 10⁶ of 293TT cells per well were plated in 100 μL of phenol red-free culture media in 96-well flat-bottom plates (3596; Costar). The cells were incubated in a 37°C tissue culture incubator for a minimum of 2 hr prior to the addition of serially diluted PsV particles. Two-fold serial dilutions of PsV particles were performed in a 2 mL round-bottom 96-well plate (40002–014; VWR) using phenol-red free culture media. From this 2 mL 96-well round-bottom plate, 100 μL of the serially diluted PsV particles was transferred to a regular round-bottom 96-well plate (3788; Costar), and 25 μL of phenol red-free culture media was added. The serially diluted PsV particles were incubated at 4°C for 1 hr to parallel the procedure for the 293TT cell-based neutralization assay. After the 1-hr incubation, 100 μL of the diluted PsV particle solutions were transferred to the pre-plated 293TT cells. At 48 hr post-infection, 50 μL of supernatants were collected from each well, and the plates were cultured for an additional 24 hr. At 72 hr, an additional 100 μL of the supernatants were harvested. The collected supernatants were stored at −80°C for a minimum of 1 day until SEAP assays were performed.

To perform infectivity assay using different numbers of 293TT cells, the assays were carried out as described above using 2 dilution factors of PsV16 (1:100,000 or 1:600,000) or PsV18 (1:10,000 or 1:60,000) particles to infect different numbers of 293TT cells. These PsV dilutions factors were chosen based on the results from the PsV titration study, and these selected dilution factors were projected to provide 100,000 to 300,000 RLUs after a 72-hr incubation with 0.03 × 10⁶ cells/well. Two-fold serial dilutions of 293TT cells were performed starting from 0.3 × 10⁶ cells to 0.002 × 10⁶ cells per well. The supernatants were collected after 48 and 72 hr of incubation at 37°C, and stored at −80°C until SEAP assays were performed.

Information on the study population and the assays

The neutralization/SEAP assays were conducted as part of the Uganda Immunogenicity Study,³⁶ which was a follow-up vaccination study of Cervarix® among 10–11 y old girls who were vaccinated as part of a government-run HPV vaccination demonstration program in Uganda. The program was designed to provide the recommended 3 doses of the vaccine; however, some of the participants received less than 3 doses because of absenteeism from school by these girls.³⁴ These serum samples were collected more than 24 mo after the receipt of the last dose of the vaccine.

In the Uganda Immunogenicity Study,³⁶ our laboratory performed the anti-HPV16 and HPV18 ELISAs, and 293TT cell-based neutralization/SEAP assays on randomly selected subset of the samples from each vaccine dose group. This enabled us to examine the titers obtained from the ELISAs correlated with those from the SEAP assays, as we have done in our previous study.²⁸ In the current study, these results were used to evaluate 72 samples, and the study protocol was approved by the National Cancer Institute Special Studies Institutional Review Board and by PATH Research Ethics Committee, Seattle, Washington, USA.

Neutralization assay and SEAP assay

For the neutralization assay, 0.03 × 10⁶ 293TT cells were plated and incubated for a minimum of 2 hr at 37˚C. Four-fold serial dilutions of serum samples were performed starting from 1:10 to 1:2,621,440, and incubated with PsV16 (1:350,000 dilution) or PsV18 (1:40,000 dilution) for 1 hr at 4°C in a final volume of 125 μL. After the incubation, 100 μL of the serum sample/PsV mixture was transferred to the pre-plated 293TT cells, and incubated for 72 hr at 37°C. The cell-culture supernatants were harvested and stored at −80°C until SEAP assays were performed.

The level of SEAP in 293TT cell-culture supernatants was measured using Great EscAPe^™ SEAP Chemiluminescence Kit 2.0 (GE) from Clontech (631738) and ZiVa® Ultra-Sensitive SEAP Detection Kit (ZiVa) from Jaden Bioscience, Inc. (CM025). The two assays were conducted simultaneously to test the same samples using white OptiPlate-96 plates (6005290; Perkin-Elmer). For GE, 25 μL of supernatant was incubated with 75 μL of 1X Dilution Buffer, and incubated at 65°C in an oven (Heratherm OMH100, Thermo Scientific) for 45 min.^27,28 After equilibrating the plates to room temperature, 100 μL of substrate was added and incubated in dark at room temperature for 20 min before measuring SEAP activity. For ZiVa, 5 μL of supernatant samples were incubated with 75 μL of SEAP Sample Preparation Solution at 65°C in the oven for 45 min. After equilibrating the plates to room temperature, 25 μL of substrate was added and incubated in the dark at room temperature for 30 min. For both kits, the activity of SEAP was measured using a plate reader (SpectraMax M5; Molecular Devices) using 0.2 s integration time. The neutralization titers were calculated by linear interpolation, and defined as the reciprocal of the dilution that showed 50% reduction in secreted alkaline phosphatase activity compared to control wells (virus-only without serum sample). Neutralization titers below the starting dilution of 1:10 were arbitrarily assigned a value of 5.

Statistical analysis

Spearman rank correlation was performed to compare anti-HPV titers in serum samples of Cervarix®-immunized individuals. GraphPad Prism 4 (GraphPad Software) was used for both analyses. Coefficient of variation (CV) for duplicate wells were calculated as follows. The replicate titer values were log transformed, and then Proc varcomp (SAS Institute, Inc.) was used to estimate the variance component (repeat measures). CVs for the duplicate wells for anti-HPV16 and -HPV18 were 24.9% and 21.9%, respectively, for GE, and were 33.0% and 29.2% for ZiVa, which were comparable to previous findings using GE.²⁹

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.

Author Contributions

All authors have reviewed the data and read the manuscript. LAP was responsible for the oversight of the study. TJK and GS designed and performed the experiments. MS is a co-PI of the Uganda Immunogenicity Study and have provided us with access to vaccinated samples. KM and LAP formulated the manuscript. TJK, KM, and LAP wrote the manuscript. TJK and KM contributed equally to this work.

Acknowledgments

We would like to extend our thanks and appreciations to PATH for making it possible for us to test the supernatant samples from the Uganda Immunogenicity Study. We would also like to thank Dr. D. Scott LaMontagne, principal investigator for the Uganda Immunogenicity Study, for his suggestions and comments on the manuscript. We would also like to thank Yuanji Pan for his technical advice on pseudovirion production.

Funding

This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US. Government.