Characteristic systemic cytokine responses in children with human bocavirus-positive lower respiratory tract infection



This article is corrected by:

  1. Errata: Corrigendum Volume 59, Issue 10, 641, Article first published online: 16 October 2015


To investigate systemic cytokine responses in human bocavirus (HBoV)-associated lower respiratory tract infection, serum cytokine profiles were analyzed in HBoV positive-children (n = 14) using multiplex immunoassay. Concentrations of TNF-α, IL-2, IL-5 and IL-8 on admission were significantly different from those of respiratory syncytial virus-positive children (n = 28). This unique cytokine response might partly explain some characteristic clinical features of HBoV-associated respiratory infection.

List of Abbreviations

human bocavirus


lower respiratory tract infection


respiratory syncytial virus

Since HBoV was detected in respiratory tract samples from Swedish infants and children with LRTI [1], numerous epidemiological studies worldwide have shown that HBoV has been distributed throughout every continent with varying incidence rates and seasonality. In addition, HBoV-positive patients have been reported to manifest several unique clinical characteristics [2].

Few studies have addressed HBoV-specific immune responses. Recently, an in vitro study showed that, on stimulation with bocavirus VP2 virus-like particles, CD4+ T cells secrete INF-γ and may play a major role in protection against this disease [3]. In another study, analysis of nasopharyngeal aspirates from HBoV-positive patients showed that Th1 and Th2 cytokines increase without Th2 polarization in children with HBoV-positive bronchiolitis [4]. Nevertheless, the pathophysiology of HBoV infection remains unclear.

To investigate the pathogenesis of HBoV-associated LRTI further, we measured 17 serum cytokines in 14 HBoV-positive children during March 2007–December 2010. All patients were previously healthy children, aged less than 2 years (median 15 months, range 8–23 months, male 57%), with proven HBoV-positive LRTI that required hospitalization on the third day of illness (median). HBoV genome was detected in nasopharyngeal swabs by PCR [2]. Using RT-PCR, no viral genomes of RSV, enterovirus, human metapneumovirus, parainfluenzavirus or adenovirus were detected in nasopharyngeal swabs [2]. Four patients were diagnosed as having pneumonia; the others had bronchitis. During this study period, 28 children aged less than two years (median 6.5 months, range 0–24 months old, male 54%), with RSV-single positive by RT-PCR were also evaluated. Seven of these patients were diagnosed as having pneumonia (one with apnea); the others had bronchitis. Serum samples were taken on the day of hospitalization (median Day 3 of illness). Days of hospital stay were similar in the two groups. No patient was diagnosed as having bronchial asthma, food allergy or atopic dermatitis on the date of admission. Sera from 14 healthy children aged less than two years (median 12 months, range 4–21 months old, male 57%) were used as controls. This study was approved by the Ethical Committee of Tokyo Women's Medical University (#1492). Informed consent was obtained from the patients’ guardians.

Multiple cytokine measurements were taken (Bio-Plex Pro; Bio-Rad Laboratories, Hercules, CA, USA). Measured cytokines were IL-1β, IL-2, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-12(p70), IL-13, IL-17, granulocyte-colony stimulating factor, INF-γ, monocyte chemoattractant protein-1, TNF-α and vascular endothelial growth factor (Table 1). Sera were stocked at −80 °C until assay. All samples from all three groups were measured on the same plate at the same time. Differences were evaluated statistically using software (JMP ver. 9; IBM). Serum values less than the detectable value were set to the minimum cut-off level of each cytokine (Table 1). Comparison between groups was made using the Mann–Whitney U-test. A P value less than 0.05 was regarded as statistically significant.

Table 1. Profiles of serum cytokines in the study subgroups
 Lower limitControlHBoVRSV 
 Quantified pg/mLn = 14 median (range) pg/mLn = 14 median (range) pg/mLn = 28 median (range) pg/mLP-valuea
  1. G-CSF, granulocyte-colony stimulating factor; VEGF, vascular endothelial growth factor.
  2. a, P-values evaluated between HBoV and RSV groups.
IL-1β0.82.0 (0.8–2.4)3.3 (1.7–7.1)2.7 (1.5–5.2)NS
IL-21.61.6 (1.6–19.3)10.1 (1.6–48.3)1.6 (1.6–30.7)0.009
IL-40.18.6 (4.6–10.4)9.4 (4.6–10.5)9.7 (4.7–11.5)NS
IL-50.72.2 (1.0–3.4)6.4 (0.7–49.1)3.0 (0.7–6.8)0.003
IL-60.66.4 (3.8–8.2)41.7 (15.9–155)24.1 (6.0–228)NS
IL-70.710.3 (5.3–15.4)17.9 (5.2–27.0)14.6 (3.2–27.7)NS
IL-80.612.5 (8.6–20.8)23.6 (17.0–121)40.2 (13.3–103)0.02
IL-92.520.7 (14.3–135)40.0 (21.1–141)30.1 (4.6–66.1)NS
IL-100.64.4 (1.9–6.7)14.8 (5.0–27.1)15.2 (6.2–102)NS
IL-12(p70)3.530.0 (9.2–88.9)48.1 (13.8–131)52.0 (4.3–149)NS
IL-130.95.5 (2.3–9.1)8.0 (5.0–19.2)8.1 (1.9–21.5)NS
IL-178.749.6 (14.1–66.4)59.8 (21.0–142)59.4 (20.1–84.6)NS
G-CSF2.823.7 (17.9–29.2)59.5 (24.9–120)55.9 (23.2–194)NS
IFNγ36.863.7 (37.3–151)140 (36.8–397)97.0 (36.8–259)NS
MCP-12.019.0 (9.9–28.9)32.8 (9.5–78.5)38.1 (11.5–153)NS
TNFα7.67.6 (7.6–16.7)51.2 (7.6–189)22.1 (7.6–60.1)0.001
VEGF0.888.3 (27.8–219)159 (38.5–710)168 (18.9–815)NS

Most cytokine concentrations were significantly higher in HBoV-positive children than in controls, the exceptions being IL-4 and IL-12(p70). Serum cytokines that were significantly different between HBoV-positive patients and RSV-positive patients were TNF-α (P < 0.001), IL-5 (P = 0.003), IL-2 (P = 0.009) and IL-8 (P = 0.02) (Table 1; Fig. 1). Serum concentrations of TNF-α, IL-5 and IL-2 were significantly higher in HBoV-positive children; however, concentrations of IL-8 were significantly lower in HBoV-positive than in RSV-positive children (Table 1; Fig. 1).

Figure 1.

Concentrations of four cytokines that differ significantly between HBoV and RSV group.

The bottom of each box represents the 25th percentile, the middle line the 50th percentile and the top the 75th percentile. Vertical lines represent 10th and 90th percentiles. Closed squares represent means of the data. HBoV, data of HBoV-positive patients (n = 14); RSV, data of RSV-positive patients (n = 28); control, data of controls (n = 14). P-values were evaluated between the HBoV and RSV groups. *, P = 0.009; **, P = 0.003; †, P = 0.02; ‡, P = 0.001.

Immune responses of RSV patients have been investigated by analysis of nasal fluid/bronchoalveolar lavage fluids, serum and circulating mononuclear cells (5–8). A Th1/Th2 cytokine balance toward increased Th2 cytokines has been described. However, there are very few such studies of HBoV-positive patients [3, 4]. In this study, we found high concentrations of multiple cytokines in HBoV-positive children in the acute stage of respiratory infection. In addition, concentrations of IL-2, IL-5 and TNF-α were significantly higher than those in RSV-positive patients.

Serum concentrations of TNF-α were most strikingly different between the HBoV and RSV groups. They are reportedly high in nasal fluid of HBoV-positive patients [4]. TNF-α reduces the effectiveness of the host immune response by impairing alveolar macrophage TNF production, thereby reducing neutrophil recruitment and phagocytosis [9]. Bauer et al. reported that concentrations of serum TNF-α correlate strongly with the degree of lung injury [10]. Yende etal. reported that preinfection systemic concentrations of TNF-α correlate with risk of community-acquired pneumonia requiring hospitalization [11]. TNF-α might be an important cytokine in the pathogenesis of HBoV-associated LRTI, the clinical manifestations of which differ from those of RSV-associated LRTI [2].

Increased serum IL-5 concentrations are reportedly present in children with several virus-associated LRTIs, including those associated with rhinovirus [12], RSV [12] and H1N1 influenza [13]. IL-2 exerts a strikingly pleiotropic array of effects on numerous target cells, various activities having been described [14]. Okamoto etal. showed that IL-2 in synergy with IL-18 induces lethal lung injury in mice [15].

We found no skewing of the Th1/Th2 balance to Th2 > Th1 in the HBoV group. It is interesting that IL-8 serum concentrations were even higher in RSV-positive than in HBoV-positive children. Regarding chemokines in RSV infection, some earlier studies have reported that RSV up-regulates chemokines in vivo and in vitro [5, 16].

Differences in ages of HBoV-positive and RSV-positive patients might have affected our results. Our study sample size was too small to provide conclusive results: additional data must be accumulated to confirm our findings. Additionally, this study did not survey all respiratory viruses, influenza virus being one of such viruses not studied.

The unique pattern of systemic cytokine response in HBoV-positive LRTIs described here may partly explain some characteristic clinical features of patients with HBoV-associated respiratory infection.


This study was supported by grants from the Japanese Society for Pediatric Infectious Diseases (H.H.) and the Chiba Serum Institute Memorial Fund (H.H.).


The authors declare that they have no conflict of interest related to this study.