Seroprevalence of Neutralizing Antibodies against Six Human Adenovirus Types Indicates the Low Level of Herd Immunity in Young Children from Guangzhou, China

Human adenoviruses (HAdVs) are non-enveloped, double-stranded DNA viruses and are divided into seven species (A–G). To date, approximately 100 types have been recognized according to genomic sequences (https://hadvwg.gmu.edu/) (Seto et al. 2011; Dehghan et al. 2019). HAdVs are highly contagious pathogens that exhibit a variety of tissue tropisms, and cause many diseases, such as acute respiratory disease (ARD), gastroenteritis, cystitis, keratoconjunctivitis, carditis, meningoencephalitis, depending on the infection type. The susceptible populations of HAdV infection include infants, immunocompromised patients (Sandkovsky et al. 2014).

Many HAdV types are relevant to ARD, including HAdV species B (type 3, 7, 14, 55, 11, 16, 21, 50), species C (type 1, 2, 5, 6, 57), and species E (type 4) (Lynch and Kajon 2016). Of these HAdVs, HAdV3, -4, -7, -14, and -55 have been reported to be associated with severe ARD outbreaks in military and civilian populations, and have caused severe and even fatal infections in both children and adults (Chen et al. 2015; Lin et al. 2017; Kajon et al. 2018; Li et al. 2018; Yao et al. 2019; Bautista-Gogel et al. 2020). HAdV-3 and -7 are the most common types in pediatric patients with ARD, and HAdV-7 is more likely to cause life-threatening pneumonia (Deng et al. 2013; Lin et al. 2017; Fu et al. 2019). Recently, a re-emergent genome type HAdV-7d associated with fatal pneumonia has been reported in Asia and the United States (Scott et al. 2016; Yu et al. 2016). HAdV-4 and -7 are the most prevalent types of HAdVs that have caused febrile ARD outbreaks in the military (Kajon et al. 2007; Cheng et al. 2016; Zhang et al. 2019). HAdV-14 and -55 are two re-emerging types that have caused numerous outbreaks among both civilian and military populations since 2006. HAdV-14 infection outbreaks have been reported in America (Kajon et al. 2010), Europe (Carr et al. 2011), and China (Huang et al. 2013). HAdV-55 is an inter-typic recombinant of HAdV-11 and HAdV-14, and HAdV-11 is the predominant type observed in patients with hemorrhagic cystitis (Numazaki et al. 1973). HAdV-55 has become a common pathogen causing severe pneumonia in northern China (Deng et al. 2013; Lu et al. 2014; Tan et al. 2016).

Vaccination may be the most effective way to prevent viral infection and establish herd immunity. However, there is no vaccine currently available for use in children and adults in the general population, although an oral vaccine comprising live HAdV-4 and -7 has been approved and used in the United States military for 40 years (Chen and Tian 2018; Gray 2020). This prompted us to develop a licensed vaccine to prevent HAdV infection outbreaks in susceptible populations (Qiu et al. 2012; Tian et al. 2014, 2015, 2018, 2019; Chen and Tian 2018; Liu et al. 2018). Therefore, it is important to investigate the predominant epidemic types and pre-existing immunity levels among populations against these types. Herd immunity is defined as the proportion of persons with immunity in a given population. The vaccination objective for public health is to increase the level of herd immunity to that affording indirect protection to unimmunized persons in a given population. The level of herd immunity can be assessed by antibody surveys (sero-epidemiology). Although many recent surveys have investigated the epidemiology of HAdV types in both children and adults with ARD, the pre-existing antibody levels have rarely been reported. HAdV infection induces type-specific, long-term neutralizing antibody (NAb) responses. NAbs against HAdV is indicative of protective antibody levels and previous infection among populations. It is helpful to investigate the prevalence of NAbs against HAdV-3, -4, -7, -14, -55, and -11 in the general population for understanding human immunity against these HAdV types and for guiding vaccine development. Although we previously investigated the seroprevalence of NAbs against HAdV-3, -4, -7, -14, and -55 in healthy adults in southern China (Tian et al. 2016; Zheng et al. 2017; Ye et al. 2018), the prevalence of NAbs in children remains unclear.

In this study, we aimed to investigate the seroprevalence of NAbs against HAdV-3, -4, -7, -14, -55, and -11 in healthy populations aged between 0 months and 49 years in Guangzhou, southern China.

Serum samples were randomly collected from 228 healthy children who received health examinations from February to June 2017 at Guangzhou Women and Children's Medical Center (n = 228, 82.01%), and 50 healthy adult blood donors at Guangzhou Blood Center from February to June 2017 (n = 50, 17.99%) in Guangzhou, southern China. The age of donors ranged from 0 months to 49 years, which were divided into six groups: 0–5 and 6–11 months old, 1–2, 3–5, 6–17, and 18–49 years old. There were 37–58 donors in each age group. The data were analyzed anonymously. However, no other detail about the donors was available.

Wild-type HAdV-11 Slobitski strain (GenBank No. AF532578.1) was purchased from American type culture collection (ATCC) and kept in State Key Laboratory of Respiratory Disease. HAdV-3 GZ01 (GenBank No. DQ099432), HAdV-4 GZ01 (GenBank No. KF006344.1), HAdV-7 GZ08 (GenBank No. GQ478341.1), and HAdV-55 Shanxi-Y16 (GenBank No. KF911353.1) strains were maintained at State Key Laboratory of Respiratory Disease. HAdV-14 GZ01 strain (GenBank No. JQ824845.1) was kindly provided by Prof. Qiwei Zhang (Southern Medical University, China). All HAdVs were cultured in HEp-2 or AD293 cells. The cells, maintained in our laboratory, were cultured in Dulbecco's modified Eagle's medium (DMEM, Gibco, USA), supplemented with penicillin (100 IU/mL), streptomycin (100 μg/mL), and 10% fetal bovine serum (Gibco, USA).

The NAb titers against the human adenoviruses were calculated using standard in vitro MN assays. In brief, the AD293 cells were seeded into 96-well plates at the density of 2 × 10⁴ cells per well and cultured overnight. Further, human sera were heated at 56 ℃ for 30 min and diluted with DMEM at 1:18. Then the sera were serially two-fold diluted (from 36- to 1152-fold), and incubated with 100 TCID₅₀ (50% tissue culture infectious dose) viruses at 37 ℃ for 1 h. Then, the mixtures were added to the 96-well plates and incubated at 37 ℃ for 48 h. The neutralization titers from triplicate wells were defined as the highest dilution of sera that inhibited the adenovirus growth without visible cytopathic effect.

The statistical significance of seroprevalence between different groups was tested using the chi-squared test, and Fisher's exact method when appropriate. The correlation between serum titer levels of HAdVs and age groups was assessed using Goodman–Kruskal Gamma method. All statistical analyses were computed with SPSS software (ver. 17.0) and P values of < 0.05 were considered significant.

The titers of NAbs against several HAdV types were measured in the sera from 278 healthy donors aged between 0 months and 49 years. As shown in Table 1, among the 278 serum samples, 122 (43.89%) were positive for at least one of the six HAdV types. We observed that a high proportion of samples (34.21%) from age group of 0–5-months-old had NAbs for at least one of the six HAdV types, which was higher than that in age groups of 6–11 months old (13.51%) and 1–2 years old (15.52%). In donors aged between 6 months and 2 years, the percentage of samples with NAbs was very low. Above the age of 1 year, the percentage of samples that was positive for HAdV NAbs (titers > 18) increased with the age of the donors (Fig. 1). The group of 18–49 years old had the highest proportion of positive samples (90%) (P < 0.001).

Figure 1.  Seroprevalence against different human adenovirus (HAdV) types in different age groups. m: month; y: year.

The seroprevalence of NAbs against different HAdV types (HAdV-3, -4, -7, -11, -14, and -55) in different age groups was investigated to illustrate the impact of age on HAdV infection. As shown in Table 1 and Fig. 1, the seroprevalence of the adult group (18–49 years old) was the highest among all HAdV types. In adults aged between 18 and 49 years, HAdV-3 (78%), -4 (56%), and -7 (54%) were the most common types, followed by HAdV-14 (38.00%) and -55 (34.00%). HAdV-11 (24.00%) was the rarest type. There was a significant difference between the prevalence of different types in 6–17-year-old group and 18–49-year-old group (P < 0.001). It is noteworthy that the percentage of positive samples from donors less than 3 years old was much lower for all six HAdV types. Children under 18 years of age showed significantly increased seropositive rates against HAdV-3 from 12.07%, 33.96% to 64.29%, HAdV-7 from 0%, 18.87% to 19.05% at ages of 1–3, 4–6, and 7–18 years, respectively, and the seroprevalence was very low (0% ~ 8.1%) for all other four types. These results indicated high infection rates of HAdV-3 and -7 in children, high infection rates of HAdV-4 and -7 in adults, and the low rate of herd immunity against HAdV-14, -55, and -11 in all ages.

NAbs titers against different HAdV types was performed (Table 2). The NAbs titers were divided into four levels: negative (< 18), low titer (18–144), medium titer (145–576), and high titer (> 576). As shown in Table 2, 38.13%, 18.71%, 14.39%, 7.55%, 8.99%, and 5.40% of the 278 samples were positive for HAdV-3, -7, -4, -55, -14, and -11, -respectively; and 8.63%, 6.12%, 2.52%, and 2.88% of the samples had notably high NAb titers (> 576) against HAdV-3, -4, -7, and -55, respectively. There was a significant difference in the seroprevalence against different HAdV types (P < 0.001). As shown in Fig. 2, a significantly higher proportion of NAb-positive samples had high titers of NAb (> 576) against HAdV-4 and -55 (P < 0.001). Most of the NAb-positive samples against HAdV-14 were in the group of low titers. There was no sample in the group of high titers against HAdV-11.

Figure 2.  Distribution of NAb titer levels against HAdV types in healthy populations. A NAb titer levels against human adenovirus types. B Distribution of NAb titers against HAdV types in positive samples. The NAb titers of 278 sera samples were determined using neutralization tests in vitro against six HAdV types. NAb, neutralizing antibody. ^*P < 0.05 for the difference in prevalence between titer groups.

Further comprehensive analyses were performed on the distribution of NAb titers against the six HAdV types in different age groups. As shown in Fig. 3, 0%, 2.7%, 3.4%, 9.4%, 26.2%, and 10% of samples from age groups of 0–5 and 6–11 months, 1–2, 3–5, 6–17, and 18–49 years had notably high NAb titers (> 576) against HAdV-3. Children in the age group of 6–17 years showed the highest rate of high NAb titer against HAdV-3. There was no sample with high-titer NAbs against HAdV-7 in the age groups of 0–5 and 6–11 months and 1–2 years; 3.77%, 4.76%, and 6% of samples from age groups of 3–5, 6–17, and 18–49 years had notably high NAb titers (> 576) against HAdV-7. The samples from children aged under 18 years showed only low-titer NAbs (18–144) against all other four HAdV types. In adults, the samples showed a relatively high proportion of high-titer NAbs against HAdV-4 (34%) and -55 (16%). Only 2% of samples from adults had high NAb titers against HAdV-14. There was no sample with high-titer NAbs against HAdV-11 in all age groups.

Figure 3.  Neutralizing antibody titers against human adenovirus (HAdV) types in different age groups. m: month; y: year.

Many serum samples were positive to two (total 27 samples) or more (total 35 samples) HAdV types, of which four samples (all from the age group of 18–49 years old) were positive to all six HAdV types. The HAdV-7-positive rate was significantly higher in HAdV-3-positive samples than in HAdV-3-negative ones and vice versa; The HAdV-7-positive rate was significantly higher in HAdV-4-positive samples than in HAdV-4-negative ones and vice versa (data not shown). These results may be attributed to the cross-reaction of one HAdV type NAb with other HAdV types, and another possibility is that some individuals might be susceptible to infection by multiple HAdV types and thus generated NAbs to them.

Then we analyzed the frequency of double-seropositive donors and single-seropositive donors against HAdV-55 and HAdV-11 in detail. The HAdV-55-positive rate was significantly higher in HAdV-11-positive samples than in HAdV-11-negative ones; Similarly, HAdV-11-positive rate was significantly higher in HAdV-55-positive samples than in HAdV-55-negative ones (χ2-test, P < 0.0001; Table 3). Similar trends were detected for HAdV-55 and HAdV-14, or HAdV-11 and HAdV-14 (data not shown). The numbers of single-positive, double-positive and tri-positive donors against HAdV-55, HAdV-11, and HAdV-14 were shown in Fig. 4A. Of total 278 samples, 6 samples were tri-positive to HAdV-55, HAdV-11, and HAdV-14, all of which had high titers of HAdV-55 NAb (Fig. 4B). In contrast, only one sample in double-negative against HAdV-14 and HAdV-11 had high titer of HAdV-55 NAb (χ2-test, P < 0.01; Fig. 4B). Further, in HAdV-11-positive samples compared to HAdV-11-negative ones, the frequency of samples with high titers of HAdV-55 NAb was much higher (χ2-test, P < 0.01; Fig. 4C). Similarly, many more donors with high titers of HAdV-55 NAb were HAdV-14-positive (χ2-test, P < 0.05; Fig. 4D). However, the frequency of samples with high or moderate levels of HAdV-14 or HAdV-11 NAb was not significant different in HAdV-55-positive compared to HAdV-55-negative samples (Fig. 4C and 4D). Totally, HAdV-55-seropositive samples tended to be HAdV-11- or HAdV-14-seropositive.

Figure 4.  Profiling of the seropositive rates and NAb titers in single-, double- and tri-positive cases for HAdV-55, HAdV-11 and HAdV-14. A The case numbers of single-, double- and tri-positive cases for HAdV-55, HAdV-11 and HAdV-14 NAbs. B The distributions of HAdV-55-positive donors with different NAb titers in both HAdV-14- and HAdV-11 positive or negative groups. C The distributions of HAdV-55-positive cases with different NAb titers in HAdV-11-positive or negative groups, and that of HAdV-11-positive donors in HAdV-55-positive or negative groups. D The distributions of HAdV-55-positive cases with different NAb titers in HAdV-14-positive or negative groups, and that of HAdV-14-positive donors in HAdV-55-positive or negative groups. The difference between the groups was analyzed by χ2-test.

In the past decade, HAdV-3 and -7 have been the two predominant HAdV types causing ARD in children in southern China, sometimes leading to outbreaks. The most common type was HAdV-3 for many years (Han et al. 2013; Chen et al. 2016; Xie et al. 2019; Yao et al. 2019). Most patients are younger than 5 years. Here we found the seroprevalence was very low in the age groups of 6–11 months and 1–2 years for all six HAdV types (Fig. 1). However, a notable proportion of samples from age groups of 0–5 and 6–11 months had NAbs as shown in Fig. 1 and Table 1, which should be inherit immunity (IgG) from mother. Therefore, children under 1 year old were rarely infected by HAdVs. Meanwhile, we observed notably high NAb titers (> 576) against HAdV-3 in 2.7% and 3.4% samples from age groups of 6–11-months and 1–2-years, respectively, which indicated that HAdV-3 infection was rare but do exist in these young children (Fig. 3). The seroprevalence against HAdV-3 and -7 increased significantly in age groups of 3–5 and 6–17 years (Fig. 1). For all other four types, the seroprevalence in the children groups was very low (0% ~ 8.1%). These results suggest that 3–5-year-old children are a high-risk population for HAdV-3 and HAdV-7 infections, and a novel vaccine against these two types, targeting pre-nursery children under the age of 3 years should be developed.

Since 2018, HAdV-7 has replaced HAdV-3 to be the most predominant type causing ARD in children in Guangzhou and has caused many severe infections (our unpublished data). This study enrolled 278 healthy donors in 2017, of which 228 were children aged between several days and 17 years. We found that children of 3–5 and 6–17 years of age had higher seropositive rates against HAdV-3 (33.96% and 64.29%, respectively) than that against HAdV-7 (18.87% and 19.05%, respectively). We also found that children of 3–5 and 6–17 years of age had a high rate of high-titer NAbs against HAdV-3 (9.4% and 26.2%, respectively). However, only 4.76% and 6% samples had notably high NAb titers against HAdV-7 from age groups of 6–17 and 18–49 years, respectively. Thus, nursery- and school-age children are high-risk groups. The lower seroprevalence and lower rate of high-titer NAbs against HAdV-7 in children reflects the lower-level of pre-existing immunity to HAdV-7, which may contribute to the change of HAdV-7 replacing HAdV-3 to be the most predominant type.

HAdV-4 and -7 caused febrile ARD outbreaks in the military, and are also circulating in civilian population. Patients infected with HAdV-4 and HAdV-7 were at increased risk for severe disease (Coleman et al. 2020). HAdV-14 and -55 are two re-emerging types. HAdV-14 has caused outbreaks among both civilian and military populations in America, Europe, and China since 2006, while HAdV-55 has become a common pathogen in northern China. However, these two types were never reported in Guangzhou, and rarely reported in southern China (Chen et al. 2016). In this study, a sizable proportion of the samples in adults were positive for NAbs against HAdV-4 (56%), -55 (34.00%), -14 (38.00%), and -11 (24.00%). We also found that a relatively high proportion of adults had high titers of NAbs for HAdV-4 (34%) and -55 (16%) while only 2% of adults had high NAb titers against HAdV-14. However, there was no sample with NAbs against HAdV-55 in children aged 1 year to 17 years, and no sample with high NAb titers against HAdV-4, -14, and -11 in children. These results demonstrated that HAdV-4, -55, and -14 were already prevalent in Guangzhou area, mainly in adults, but few in children before 2017. Similarly, in our previous study with sera collected in Dongguan City, a city near Guangzhou, in 2014, the percentage of samples from age groups of 20–29, 30–39, and 40–49 years that were positive for HAdV-55 NAbs was 0%, 19.75%, and 40.625%, respectively (Tian et al. 2016). This study mainly investigated the pre-existing immunity in young children, so the donors were divided into six groups: 0–5 and 6–11 months old, 1–2, 3–5, 6–17, and 18–49 years old. There were 37–58 donors in each age group. In future work, more detailed study could be done, including more donors in the age groups of 6–17, and over 18 years old which could be divided into more groups, such as 6–8, 9–11, 12–14, 15–17, 18–21, 22–29, 30–39, 40–49, 50–59, 60–69, over 70. The proportion of samples from our donors in Guangzhou that were positive for NAbs against HAdV-55 or -14 was low in both children and adults, which suggests that there may be little protection in the Guangzhou population against HAdV-55 or -14 infection. However, the proportion of samples that were positive for NAbs against HAdV-4 or -7 was high in adults, but very low in children, which indicated the low level of protective immunity against HAdV-4 and -7 in children. More attention should be paid to these HAdVs circulating in civilian populations. HAdV-4, -7, -55, and -14 vaccinations may be urgently needed for young people under 18 years of age.

HAdV-55 is an inter-typic recombinant of HAdV-11 and HAdV-14, sharing a similar but somewhat different hexon protein with HAdV-11 and a similar fiber protein with HAdV-14 (Walsh et al. 2010). Furthermore, HAdV-11 was indirectly excluded in this study. Furthermore, HAdV-11 was indirectly excluded in this study. Here we showed HAdV-55-seropositive samples tended to be HAdV-11- or HAdV-14-seropositive (Table 3 and Fig. 4). Early studies reported fiber-specific NAbs might contribute to the cross-neutralizing against HAdV-14 and HAdV-55 of double-positive sera (Feng et al. 2018). The cross-neutralizing against HAdV-55 and HAdV-11 may be attributed to hexon, which elicited predominant NAbs in most individuals who were infected by HAdVs. We also found few individuals had high titers of NAbs against HAdV-55 but no NAb against HAdV-11 (Fig. 4C). This result may be attributed to the difference among hexon proteins of HAdV-55 and HAdV-11, or NAbs targeting other capsid antigens, such as penton base or fiber. However, these speculations should be investigated in future studies. The low seroprevalence and low NAb titers against HAdV-11 in all age groups were consistent with previous studies, which indicated the advantages of being vectors for vaccination, gene therapy, and cancer therapy (Holterman et al. 2004). HAdV-11 transduces primary cells including smooth muscle cells, synoviocytes, dendritic cells, and cardiovascular tissues with higher efficiency than HAdV-5. Furthermore, HAdV-11 vector elicited immune responses both in the presence and absence of anti-HAdV-5 immunity (Lemckert et al. 2005; Stone et al. 2005).

In summary, we investigated the seroprevalence and distribution of NAbs against HAdV-3, -4, -7, -14, -55, and -11 by age in a civilian population, including children and adults aged between 0 months and 49 years in Guangzhou. These results revealed a low level of herd immunity against HAdV-14, -55, and -11 in all age groups, and a high level of herd immunity against HAdV-3, -4, and -7 in adults. This study also indicates a high level of herd immunity against HAdV-3 in children in the age group of 7–17 years, and the low rate of herd immunity against HAdV-4 and -7 in children of all age groups. These results suggest the importance of monitoring HAdV types, and developing prophylactic vaccines for children and young adults.

This work was supported by the National Key Research and Development Program of China (2018YFC1200100); the Guangzhou Science and Technology Program Key Project, China (201803040004); the National Science and Technology Major Project of China (2017ZX10103011003, 2018ZX1010 2001); the National Natural Science Foundation of China, China (31570163); the Youth Project of State Key Laboratory of Respiratory Disease, China (SKLRD-QN-20 1713).

TG conceived and designed the experiments. FY, TG, WB, LW, XY, MC, and YP carried out the experiments. TG, FY, LK, and LX analyzed the data. TG and FY wrote the paper. RX and ZR checked and finalized the manuscript. All authors read and approved the final manuscript.

The authors declare that they have no conflict of interest.

This study was approved by the Ethics Committee of the Affiliated First Hospital of Guangzhou Medical University. Informed consent was obtained from all the participants.