Decreased hemoglobin levels in pneumonia patients associated with the ybt gene of Klebsiella pneumoniae

To the Editor: Klebsiella pneumoniae (KP) is a common opportunistic pathogen causing potentially life-threatening illness and organ failure. Surveillance data from the 2023 China Antimicrobial Surveillance Network (CHINET) revealed that the KP isolation rate from clinical respiratory specimens in China was the highest (18.1%) among all bacteria.[1] Both classical (cKP) and hypervirulent KP (hvKP) exert significant effects on the systemic inflammatory response, causing infection-related anemia and high mortality to host. The main gap in current research is that the specific virulence factors of KP that cause host infectious anemia have not been identified. This study aimed to identify the key KP virulence factors causing infectious anemia in patients with pneumonia. Patients diagnosed with KP-associated pneumonia at the Peking University Third Hospital from January 2017 to April 2023 were enrolled. The inclusion criteria were age ≥18 years, and KP isolation by culture from lower respiratory tract, blood, or urine specimens. The exclusion criteria were pre-existing primary or secondary hematological diseases, pregnancy, or patients lacking routine blood test records during the study period. Data on KP strain sequences, blood routine results (white blood cell count, red blood cell count [RBC], hemoglobin [HGB], platelet count, lymphocyte percentage and count [LYM% and LYM#, respectively], neutrophil percentage and count, monocyte percentage and count [MONO% and MONO#, respectively]), and post-infection prognosis (mortality within 14, 28, and 90 days) were collected. Bioinformatics analysis was used to identify virulence factors such as iroB, iutA, rmpA, rmpA2, and peg-344. Inclusion of any of these virulence factor genes defined the strain as hvKP, while strains without these genes were defined as cKP.[2] Independent risk factors for infection-related anemia were analyzed through group comparison, subgroup comparison, and logistic regression analysis. The study was approved by the ethics committee of Peking University Third Hospital (No. M2021545) and followed the Declaration of Helsinki. The requirement to obtain the informed consent was waived. A total of 627 patients with KP associated pneumonia were enrolled, including 447 males (71.3%) and 180 females (28.7%). Patient age ranged from 18 to 107 years. The main multi-locus sequence typing (MLST) of isolated KP strains was ST11 (138/627, 22.0%), and the main capsular serotype was KL1 (90/627, 14.4%). Gene sequencing showed the positive rate of key virulence factors was 66.2% (415/627) for ybt, 49.1% (308/627) for peg-344, 48.8% (306/627) for iutA, 48.5% (304/627) for rmpA, 46.1% (289/627) for iroB, and 45.6% (286/627) for rmpA2, respectively. More patients were infected with hvKP (362/627, 57.7%) than with cKP (265/627, 42.3%). Compared with the hvKP group, the cKP group had lower positive rate of the ybt gene (57.7% [153/265]) vs. 72.4% [262/362], P <0.001), RBC (3.35 [2.83–3.98] × 1012/L vs. 3.87 [3.18–4.42] × 1012/L, P <0.001), HGB (100.00 [86.00–118.00] g/L vs. 118.00 [98.00–136.00] g/L, P <0.001], LYM% (10.60 [6.25–16.95]% vs. 12.70 [6.40–19.25]%, P = 0.022), LYM# (0.93 [0.63–1.51] × 109/L vs. 1.15 [0.69–1.68]×109/L, P = 0.022), higher MONO% (5.50 [3.90–7.25]% vs. 5.00 [3.30–6.60]%, P = 0.004) and MONO# (0.50 [0.37–0.69] × 109/L vs. 0.46 [0.31–0.62] × 109/L, P = 0.002). After age and gender matching, significant differences remained in RBC (3.35 [2.87–3.99] × 1012/L vs. 3.75 [3.14–4.36] × 1012/L, P = 0.001) and HGB (101.00 [87.00–121.00] g/L vs. 113.00 [94.00–135.00] g/L, P <0.001). According to the blood routine reference range provided by the laboratory of Peking University Third Hospital, HGB below 130 g/L in males and 115 g/L in females is defined as anemia. In total, 440 patients in our study had anemia, including 318 males (72.3%) and 122 females (27.7%). Further analysis revealed that the group with anemia was older (79.00 [66.00–86.75] years vs. 69.00 [58.00–80.00] years, P <0.001) than the non-anemic group. The anemic group had a higher proportion of KP strains carrying ybt (69.1% [304/440] vs. 59.4% [111/187], P = 0.018) and a lower proportion of strains carrying other virulence genes (peg344 [41.8% (184/440) vs. 66.3% (124/187), P <0.001], iutA [43.2% (190/440) vs. 62.0% (116/187), P <0.001], iroB [37.5% (165/440) vs. 66.3% (124/187), P <0.001], rmpA [40.9% (180/440) vs. 66.3% (124/187), P <0.001], and rmpA2 [40.2% (177/440) vs. 58.3% (109/187), P <0.001]). According to logistic regression analysis, age (odds ratio [OR]: 1.018, 95% confidence interval [CI]: 1.006–1.030, P = 0.003) and infection with KP carrying ybt (OR: 1.564, 95% CI: 1.072–2.283, P = 0.020) were independent risk factors leading to HGB decrease. To further clarify the role of the ybt gene in KP-associated infection-related anemia, we performed two comparative analyses in the cKP and hvKP subgroups, according to whether the strains carried ybt (ybt(-)hvKP vs. ybt(+)hvKP, and ybt(-)cKP vs. ybt(+)cKP). Compared with the ybt(-)hvKP group, the ybt(+)hvKP group had lower RBC (3.77 [3.12–4.37] × 1012/L vs. 4.10 [3.48–4.46] × 1012/L, P = 0.005) and HGB (113.00 [94.00–133.00] g/L vs. 122.50 [109.25–139.00] g/L, P = 0.001). Similar results were also observed in the cKP group [Table 1]. Table 1 - Comparison of clinical features between ybt(-) group and ybt(+) group in pneumonia patients infected with cKP and hvKP. Characteristics cKP group (n = 265) Statistics P hvKP group (n = 362) Statistics P ybt (–)cKP group (n = 112) ybt (+)cKP group (n = 153) ybt (–)hvKP group (n = 100) ybt (+)hvKP group (n = 262) Age (years) 80.50 (68.00, 88.00) 83.00 (69.00, 88.00) –0.786* 0.432 67.50 (56.00, 77.00) 72.00 (62.00, 84.00) –3.101* 0.002 Male 74 (66.1) 103 (67.3) 0.045† 0.831 76 (76.0) 194 (74.0) 0.146† 0.703 Diabetes mellitus 36 (32.1) 53 (34.6) 0.181† 0.671 44 (44.0) 104 (39.7) 0.555† 0.456 Blood routine WBC (×109/L) 9.38 (7.11, 12.91) 10.35 (6.96, 12.93) –0.102* 0.919 8.75 (6.85, 12.08) 9.64 (6.97, 12.14) –0.888* 0.375 RBC (×1012/L) 3.55±0.74 3.31±0.82 2.451‡ 0.015 4.10 (3.48, 4.46) 3.77 (3.12, 4.37) –2.785* 0.005 HGB (g/L) 111.00 (89.25, 123.75) 94.00 (83.50, 114.00) –3.026* 0.002 122.50 (109.25, 139.00) 113.00 (94.00, 133.00) –3.214* 0.001 PLT (×109/L) 217.50 (146.75, 299.25) 187.00 (127.50, 260.50) –2.228* 0.026 186.00 (133.25, 248.50) 188.00 (131.75, 242.00) –0.121* 0.904 LYM (%) 11.00 (7.05, 18.28) 9.90 (5.30, 16.35) –0.983* 0.325 12.70 (7.00, 17.88) 12.90 (6.30, 20.23) –0.466* 0.641 LYM count (×109/L) 1.00 (0.64, 1.62) 0.90 (0.62, 1.41) –1.235* 0.217 1.05 (0.64, 1.48) 1.19 (0.72, 1.73) –1.079* 0.281 NE (%) 81.50 (71.65, 87.33) 81.90 (73.80, 88.40) –0.716* 0.474 81.85 (74.8, 88.55) 80.95 (69.78, 89.03) –0.515* 0.607 NE count (×109/L) 7.31 (5.35, 10.72) 7.94 (5.41, 11.12) –0.183* 0.855 7.03 (4.96, 10.08) 7.58 (5.10, 10.17) –0.612* 0.540 MONO (%) 5.55 (3.93, 6.90) 5.40 (3.75, 7.50) –0.206* 0.837 5.00 (3.20, 6.65) 5.00 (3.30, 6.60) –0.040* 0.968 MONO count (×109/L) 0.51 (0.37, 0.67) 0.49 (0.36, 0.78) –0.036* 0.972 0.45 (0.29, 0.63) 0.46 (0.31, 0.61) –0.615* 0.539 Prognosis D14 mortality 26 (23.2) 38 (24.8) 0.093† 0.761 8 (8.0) 48 (18.3) 5.895† 0.015 D28 mortality 33 (29.5) 54 (35.3) 0.997† 0.318 9 (9.0) 64 (24.4) 10.700† 0.001 D90 mortality 50 (44.6) 73 (47.7) 0.245† 0.621 13 (13.0) 81 (30.9) 12.085† 0.001 Data are presented as median (Q1, Q3), n (%), or mean ± standard deviation. *Z values; †χ2 values; ‡t values. cKP: Classical Klebsiella pneumoniae; D14/D28/D90 mortality: Mortality on day 14/28/90 after KP infection; HGB: Hemoglobin level; hvKP: Hypervirulent Klebsiella pneumoniae; LYM: Lymphocyte; MONO: Monocyte; NE: Neutrophil; PLT: Platelet count; RBC: Red blood cell count; WBC: White blood cell count. There may be inconsistencies in the number and combination of virulence genes carried by hvKP. To rule out the effect of inconsistent virulence gene profiles in hvKP, we screened patients infected with KP strains simultaneously carrying five key virulence factors (peg-344, iutA, iroB, rmpA, and rmpA2; n = 221). Patients in the ybt(+) group (n = 144) had lower RBC (3.96 [3.36–4.47] × 1012vs. 4.14 [3.61–4.46] × 1012, P = 0.331) and HGB (120.50 [103.00–137.00] g/L vs. 125.00 [112.00–139.50] g/L, P = 0.167) compared with ybt(-) group, but no statistically significant differences in RBC and HGB were observed. Thus, we speculate that in hvKP carrying these five genes, the role of the ybt gene may be counteracted by other virulence genes. The interactions between these genes and the underlying molecular mechanisms remain unclear and require further study. Yersiniabactin, a siderophore predominantly present in Yersinia species and Enterobacteriaceae, has its biosynthesis directed by the ybt gene. Upon invading the host, KP utilizes yersiniabactin, encoded by the ybt gene, to compete with host iron-binding proteins for iron, enabling the bacteria to survive and proliferate within the host.[3,4] Our research reveals that patients infected with ybt(+) KP exhibit a more pronounced decrease in HGB levels, a finding that has not been previously documented. The exact mechanism behind this phenomenon remains elusive. It may be related to yersiniabactin aiding KP in sequestering iron, which could deplete the host’s raw materials for heme synthesis and thereby lower HGB levels. Furthermore, the ybt iron uptake system and the heme receptor gene are located on the same virulence plasmid.[5] They can work together to acquire both iron and heme from the host via the TonB system. This synergistic interaction may enhance the effect of ybt(+) KP strains in reducing HGB levels. In summary, this study revealed possible association between ybt and reduced HGB during KP infection. These results provided a possible clue for further clarifying the mutual regulatory mechanism behind KP-related infectious anemia and possible treatment directions targeting this virulence factor. Acknowledgment We thank the department of laboratory medicine of Peking University Third Hospital for providing the KP strains, and thanks to the Peking University Respiratory Infections and Microbiology Youth Innovation Research Team for providing professional consultation in respiratory field. Funding This study was supported by Beijing Key Clinical Specialty Funding (No. 010071), the Clinical Cohort Construction Program of Peking University Third Hospital (No. BYSYDL2019007), Clinical Key Project of Peking University Third Hospital (No. BYSYZD2022007), and Peking University Outstanding Doctoral Student Innovation Funding (No. BMU2024BSS001). Conflicts of Interest None.