Bone Marrow Microbiome: Metagenomic Comparsion of Childhood Acute Lymphoblastic and Acute Myeloid Leukamias

Purpose: To describe and compare the pre-treatment bone marrow metagenomes of childhood acute lymphoblastic leukemia (ALL) and childhood acute myeloid leukemia (AML).

Background: Leukemia is the most common cancer in children, comprised mainly of ALL (~80%) followed by AML (~17%). The role of infections and their timing in the etiology of pre B-cell ALL's (but not AML's) is supported by a considerable body of epidemiologic evidence developed in the past 20 years (Gilham, Peto et al. 2005, Ma, Urayama et al. 2009). There are two main infection related hypotheses. Greaves postulates that delayed infection, akin to the hygiene hypothesis, increases risk of leukemia (Greaves and Alexander 1993, Greaves 1997). The second hypothesis set forth by Kinlen is not mutually exclusive and postulates the role of a transforming yet common virus where introduction of the infectious agent to previously naive communities trigger leukemia (Kinlen 1995). While many viral agents have been investigated (reviewed in (O'Connor and Boneva 2007)) no definitive infectious agent has been identified but, until recently, broad surveys of the microbiome were not possible. Recent developments in next generation sequencing and bioinformatics now allow for unbiased investigation into the totality of organisms that inhabit the human body.

Methods: Pre-treatment bone marrow specimens were collected in pre B-cell ALL and AML cases participating in the California Childhood Leukemia Study. AML's with similar age distribution to the ALL's were utilized as controls, to represent the infection history of an immune suppressed child. We interrogated the viral metagenomes of 60 ALL’s and 30 AML’s, pooled and reduced with ribosomal capture, then sequenced with high coverage next generation RNA sequencing on an Illumina HiSeq 2500. The resulting sequences were analyzed using a custom bioinformatics pipeline, DARK, to remove human sequences, to identify viral transcripts, and to categorize all transcripts present in the pre-treatment bone marrow. Bacterial metagenomes of the same subjects were analyzed using 16S rRNA amplicon sequencing and the QIIME pipeline.

Results: We analyzed over 500 million 100bp reads and categorized human, non-human and unknown sequences. We found striking differences between the viral metagenomes of childhood ALL and AML, with ALL subjects presenting with a greater viral infection load than AML’s. Furthermore, differences in the presence of transcripts aligning to both exogenous viruses and endogenous retroviruses were observed with some viruses occurring exclusively in ALL’s. We found no difference between the bacterial metagenomes of childhood ALL and AML suggesting no differential contamination occurred during collection and analysis.

Discussion: This study provides the first measurement of all identifiable infectious agents in both childhood ALL an AML. Our data suggest that bacteria do not play a role in the etiology of ALL. The greater number and frequency of viral infectious agents were observed in ALL versus AML contribute to an etiologic hypothesis of ALL involving viral infections yet require replication and additional follow up. Moreover the clinical relevance of these infections and their relation to outcome should be investigated.

References

Gilham, C., J. Peto, J. Simpson, E. Roman, T. O. Eden, M. F. Greaves, F. E. Alexander and U. Investigators (2005). "Day care in infancy and risk of childhood acute lymphoblastic leukaemia: findings from UK case-control study." BMJ 330(7503): 1294.

Greaves, M. F. (1997). "Aetiology of acute leukaemia." Lancet 349(9048): 344-349.

Greaves, M. F. and F. E. Alexander (1993). "An infectious etiology for common acute lymphoblastic leukemia in childhood?" Leukemia 7(3): 349-360.

Kinlen, L. J. (1995). "Epidemiological evidence for an infective basis in childhood leukaemia." Br J Cancer 71(1): 1-5.

Ma, X., K. Urayama, J. Chang, J. L. Wiemels and P. A. Buffler (2009). "Infection and pediatric acute lymphoblastic leukemia." Blood Cells Mol Dis 42(2): 117-120.

O'Connor, S. M. and R. S. Boneva (2007). "Infectious etiologies of childhood leukemia: plausibility and challenges to proof." Environ Health Perspect 115(1): 146-150.