Bone marrow osteoblast damage by chemotherapeutic agents

Abstract

Hematopoietic reconstitution, following bone marrow or stem cell transplantation, requires a microenvironment niche capable of supporting both immature progenitors and stem cells with the capacity to differentiate and expand. Osteoblasts comprise one important component of this niche. We determined that treatment of human primary osteoblasts (HOB) with melphalan or VP-16 resulted in increased phospho-Smad2, consistent with increased TGF-β1 activity. This increase was coincident with reduced HOB capacity to support immature B lineage cell chemotaxis and adherence. The supportive deficit was not limited to committed progenitor cells, as human embryonic stem cells (hESC) or human CD34+ bone marrow cells co-cultured with HOB pre-exposed to melphalan, VP-16 or rTGF-β1 had profiles distinct from the same populations co-cultured with untreated HOB. Functional support deficits were downstream of changes in HOB gene expression profiles following chemotherapy exposure. Melphalan and VP-16 induced damage of HOB suggests vulnerability of this critical niche to therapeutic agents frequently utilized in pre-transplant regimens and suggests that dose escalated chemotherapy may contribute to post-transplantation hematopoietic deficits by damaging structural components of this supportive niche.

Figure 1. Melphalan and VP-16 exposure of HOB increases the level of active TGF-β1.

A) HOB were exposed to melphalan [50 µg/ml] or VP-16 [50 µM] for 24 hours and real time PCR for TGF-β1 completed. B) HOB were exposed to melphalan [100 µg/ml] or VP-16 [100 µM] for 4 hours, fixed and stained for detection of phospho-Smad2 (green), as a read-out of TGF-β1 activity (DAPI stain for nuclei-blue). C) HOB were exposed to melphalan [100 µg/ml], VP-16 [100 µM] or rTGF-β1 [10 ng/ml] with or without SB 431542 [10 µM] for 4 hours and a western blot completed for p-Smad2 and total Smad2. P-Smad2 expression was normalized to total Smad2 expression. D) HOB were exposed to rTGF-β1 [10 ng/ml] for 24 hours and real time PCR completed (left). Additionally, HOB were exposed to rTGF-β1 [10 ng/ml] for 24 hours, the HOB layer was rinsed and new media was added and allowed to condition for 24 hours before being evaluated by ELISA to quantitate the amount of secreted TGF-β1 (right). (NT = No Treatment, Mel = melphalan, VP = VP-16, TGF = rTGF-β1, SB = SB 431542).

Figure 2. Melphalan or rTGF-β1 exposure diminished the ability of HOB to support CD34+ bone marrow cells.

A) To evaluate HOB support of CD34+ bone marrow cells, HOB cells were exposed to melphalan [25 µg/ml] or rTGF-β1 [10 ng/ml] for 24 hours. After the 24 hour exposure, the HOB were thoroughly rinsed and CD34+ cells (8.8×10⁵) were added in co-culture. Recombinant IL-3 (100 ng/ml) was added in all groups. CD34+ cells were collected after co-culture for 48 hours and the samples were analyzed for the expression of cell surface markers. B) HOB were exposed to melphalan [50 µg/ml] or rTGF-β1 [10 ng/ml] for 24 hours. The HOB layers were rinsed and CD34+ bone marrow cells (1.75×10⁵) were added in co-culture for 48 hours. The CD34+ cells were collected from the HOB layers, viability determined, and 1×10³ viable CD34+ bone marrow cells were added to methocult containing factors for myeloid lineage in triplicate. Colonies were counted and scored after 7 days in culture. *p<0.05 (average number of colonies, treated vs untreated), **p<0.05 (CFU-GM colonies, treated vs untreated) and #p<0.05 (CFU-M colonies, treated vs untreated).

Figure 3. Exposure of HOB to chemotherapeutic agents or rTGF-β1 diminished the ability of HOB to support human embryonic stem cells and decreased pro-B cell interaction with HOB.

A) HOB were pre-exposed to melphalan [50 µg/ml], VP-16 [50 µM] or rTGF-β1 [10 ng/ml] for 24 hours. The HOB were rinsed thoroughly and human embryonic stem cells were added in co-culture with the HOB. After 2 days of co-culture, stem cell colonies were counted, stained for Oct-4 as an indicator of pluripotent potential and B) designated as undifferentiated or differentiated based on classic morphology of well defined borders. C) To determine if exposure of HOB to chemotherapeutic agents or rTGF-β1 affected pro-B cell adherence, HOB were exposed to melphalan [50 µg/ml], VP-16 [50 µM] or rTGF-β1 [10 ng/ml] for 24 hours. Adherent layers of HOB were rinsed thoroughly and co-cultured with fluorescently labeled pro-B cells (1×10⁵). After 24 hours the media was aspirated and non-adherent pro-B cells were gently rinsed. Remaining HOB and pro-B cells adherent to the HOB layer were trypsinized and events counted on high flow rate for 30 seconds to enumerate number of pro-B cells attached to the HOB. *p<0.05.

Figure 4. Exposure of HOB to chemotherapeutic agents or rTGF-β1 decreased the expression of CXCL-12 and diminished support of pro-B cell chemotaxis.

A) HOB were exposed to melphalan [50 µg/ml], VP-16 [50 µM] or rTGF-β1 [10 ng/ml] for 24 hours. RNA was isolated and real time RT-PCR completed for CXCL-12. B) HOB were exposed to melphalan [50 µg/ml], VP-16 [50 µM] or rTGF-β1 [10 ng/ml] for 24 hours. Following exposure, cells were rinsed and new media added and conditioned for 48, 72 or 96 hours. Supernatants were collected and evaluated by CXCL-12 ELISA. C) HOB were exposed to melphalan [50 µg/ml], VP-16 [50 µM] or rTGF-β1 [10 ng/ml] for 24 hours. Following exposure, 350 ul of supernatant was placed into the bottom chamber of a transwell with or without CXCL-12 [1 ng/ml]. HOB media and CXCL-12 [100 ng/ml] were used as negative and positive controls, respectively. JM-1 progenitor cells (1.5×10⁵) were placed in the top chamber and allowed to migrate for 4 hours. After 4 hours, the cells migrating to the bottom chamber were read on high flow rate for 30 seconds on a FACSCalibur flow cytometer. (CM = conditioned media), *p<0.05 (CM compared to treatments), #p<0.05 (individual group compared to CXCL-12 addition).

Figure 5. Global changes in gene expression in HOB following direct and indirect exposure to chemotherapeutic agents or rTGF-β1.

HOB cells were exposed to melphalan [50 µg/ml], rTGF-β1 [10 ng/ml] or conditioned media from BMSC (CMM) pre-exposed to melphalan [50 µg/ml-24 hours] for 6 hours. BMSC exposed to melphalan were rinsed and fresh media was place on adherent layers to condition prior to stimulating HOB. After the 6 hour exposure, HOB RNA was isolated and microarray analysis completed to evaluate global changes in gene expression. A) The Venn diagram summarizes the number of HOB genes that changed within each group as well as the changes between the groups. B) A network diagram was generated for the intersection of all 3 treatment groups highlighting the convergence of potential pathways associated with the genes found in common between the 3 treatment groups, such as NF-κB. All genes listed were generated using a 2.5% FDR and >1.5 fold change in significance. C) A gene summary was made common genes up-regulated (4, red) or down-regulated (2, green) for the intersection of all 3 treatments (rTGF-β1: CMM: melphalan).