PIONEERS OF OSTEOBIOLOGICS · PART TWO: THE SIGNAL

The Signal

How Urist, Reddi, and Wozney Identified the Molecules That Instruct Bone to Form — and Built the First Biologic That Could Deliver Them

Bone does not simply grow. It is instructed to grow — by signals embedded in its own matrix, waiting to be released, decoded, and put to work.

Introduction

Part One of this series established that bone marrow contains a rare population of progenitor cells capable of forming bone, cartilage, and the structural scaffolding of the skeleton. But identifying those cells raised an immediate second question: what tells them to become bone in the first place?

The answer took three scientists, more than two decades of painstaking biochemistry, and the eventual arrival of recombinant DNA technology to decode. Marshall R. Urist found the phenomenon. A. Hari Reddi mapped the biological mechanism and developed key extraction and bioassay methods that made purification possible. John M. Wozney provided the molecular identity and made pharmaceutical-scale production possible. None of them could have reached clinical translation alone — each built directly on what the others established.

This is the story of how a substance present in extremely low concentrations became the first recombinant growth factor approved by the FDA for skeletal repair.

I.  Marshall R. Urist (1914–2001)

The Man Who Found the Bone Inductive Principle

The Making of an Orthopedic Surgeon

Marshall Raymond Urist was born on June 11, 1914, in Chicago, Illinois, and raised on a small farm in South Haven, Michigan. (8) He earned a bachelor's degree in chemistry from the University of Michigan in 1936, a master's degree from the University of Chicago in 1937, and his medical degree from the Johns Hopkins University School of Medicine in 1941.(8) After residency training at Johns Hopkins and Massachusetts General Hospital, Urist enlisted in the United States Army in 1943, serving as chief of orthopedics in the 22nd General Hospital in England and the 97th General Hospital in Germany. When he returned from Germany in 1946, he was assigned to the Surgeon General's office to document advances in the treatment of open fractures — work that gave him an early, systematic view of what bone could and could not do on its own.(8)

He joined the UCLA School of Medicine faculty in 1954, a tenure that would last 46 years. He directed the bone research laboratory at UCLA, served as editor-in-chief of Clinical Orthopaedics and Related Research for 27 years, and published more than 415 papers.(8) His breadth — surgeon, researcher, editor, teacher — was unusual, and it shaped how he approached a problem that had fascinated him since medical school.

The Question That Drove Three Decades of Work

As a young physician, Urist had read Leriche and Policard's description of a hypothetical "juice of stonemaking" — the idea that something intrinsic to bone tissue was responsible for its regenerative capacity.(8) The concept had roots going back further: Charles Huggins had demonstrated that certain transplanted tissues could induce ectopic bone formation, establishing the broader principle that bone induction could occur outside the native skeleton. Huggins would win the Nobel Prize in 1966 for related cancer research. (11)

What nobody had established was what the inducing substance actually was, where it resided, and whether it could be isolated from bone itself. That was Urist's question — and it would take him thirty years to answer it fully.

"Bone Formation by Autoinduction" — 1965

Working in his UCLA laboratory over the course of nearly two decades, Urist pursued bone induction through a methodical series of implantation experiments. He took segments of cortical bone, removed the cells and mineral through demineralization and lyophilization, and implanted the remaining organic matrix into the muscle pouches of rabbits. (1)

What he observed was repeatable and striking: the demineralized matrix induced new bone formation at ectopic sites with no contribution from native skeletal tissue. (1) The bone matrix itself appeared to carry a bone-forming inductive property. Urist named this phenomenon "bone formation by autoinduction" and published the findings in Science on November 12, 1965. The paper described the active agent as a "bone morphogenetic property" residing in the organic matrix, and initiated a decades-long search to identify and isolate it. (1)

The 1965 paper was not immediately embraced. The mechanism Urist proposed — that dead, acellular matrix could instruct living host cells to form new bone — ran against prevailing assumptions about how tissue regeneration worked. The paper accumulated slowly at first, but its influence proved durable. By the late 1990’s, the paper had been recognized as a landmark contribution to modern bone biology.

New-bone formation by autoinduction in which both the inductor cells and the induced cells are derived from ingrowing cells of the host bed... Differentiation of the osteoprogenitor cell is elicited by local alterations in cell metabolic cycles that are as yet uncharacterized.

Three Decades of Isolation Work

Urist spent the years following 1965 attempting to isolate and characterize the bone-forming substance he had identified. In 1971, he and Strates formally coined the term "bone morphogenetic protein" — BMP — in the Journal of Dental Research, marking the first use of the name that would come to define an entire class of clinical biologics.(2) Many of the BMPs later identified, including BMP-2, BMP-3 and BMP-4, were found to be members of the TGF-β superfamily.

The isolation work was slow. BMP existed in bone matrix in extraordinarily small quantities — often parts per billion — and the demineralized matrix from which it had to be extracted was largely insoluble, resisting standard protein purification techniques.(10) Urist's group produced partial purifications and demonstrated clinical activity in non-union fractures and segmental defects, but a homogeneous, fully characterized BMP protein remained out of reach throughout his career. The substance was real; it was demonstrably active; it could not yet be reliably produced.

That barrier — insolubility — would be broken not in a clinical laboratory but in a biochemistry one, by a scientist who approached the problem from the mechanism rather than the molecule.

II.  A. Hari Reddi, PhD

The Man Who Mapped the Cascade — and Made Purification Possible

From the University of Chicago to the NIH

A. Hari Reddi was born on October 20, 1942. His early research career took him through the University of Chicago and eventually to the National Institutes of Health, where he built one of the most productive bone biology laboratories of the 20th century.(11) In 1997, he was recruited to the University of California, Davis, as the inaugural holder of the Lawrence J. Ellison Endowed Chair in Musculoskeletal Molecular Biology — a position created specifically to house his work.(10) He was the first recipient of the Marshall Urist Award from the Orthopedic Research Society, also in 1997 — an award named for the man whose discovery he had done more than almost anyone else to explain.(10)

Mapping the Cascade — What Actually Happens When Bone Forms

Reddi's most consequential early contribution was conceptual. Where Urist had identified that something in bone matrix induced osteogenesis, Reddi asked what the sequence of biological events actually looked like — step by step, cell by cell, day by day.

Working with Charles Huggins at the University of Chicago, Reddi published in 1972 a detailed account of the biochemical sequences involved in the transformation of fibroblasts into bone-forming cells in response to demineralized bone matrix implants.(3) The picture that emerged was not a simple switch — it was a cascade. Bone induction, Reddi showed, proceeded through a defined, multistep sequence: recruitment of progenitor cells to the implant site, proliferation, differentiation into cartilage, hypertrophy of that cartilage, vascular invasion, and finally replacement with bone and bone marrow.(12)

This cascade model had important implications. It meant that bone induction was not a single molecular event but a coordinated biological process, and that disrupting any step — inadequate progenitor recruitment, insufficient carrier, wrong dose — could abort the sequence before bone ever formed. The model foreshadowed many of the delivery problems later observed clinically with supraphysiological BMP use that would not become fully apparent for another three decades.

The Dissociative Extraction Breakthrough — 1981

Urist's isolation work had stalled against the insolubility of demineralized bone matrix. In 1981, Sampath and Reddi published a solution.(4) Using chaotropic agents — guanidinium hydrochloride and urea — they dissociatively extracted the soluble protein fraction from demineralized bone matrix, separating it cleanly from the insoluble collagenous residue.

The critical finding was this: neither fraction alone induced bone. The soluble proteins without the collagenous matrix produced no osteogenesis. The matrix without the proteins produced no osteogenesis. Only when the two were reconstituted together — soluble signal plus insoluble scaffold — did bone induction occur.(4)

This two-component principle became foundational to modern biologic/carrier combination products. The finding that a molecular signal requires a physical scaffold to function in vivo is not a packaging or delivery detail — it is a biological requirement, built into the mechanism of bone induction itself. Ignoring it, as clinical experience would later show, has consequences.

Reddi's group also developed the rat subcutaneous bioassay — a standardized in vivo test for osteoinductive activity that gave the field a reproducible, quantifiable measure for the first time.(12) That assay became the essential tool for the purification work that followed, providing the readout that guided every subsequent step toward isolating the active proteins.

Osteogenin and the BMP Family

Through the early and mid-1980s, Reddi's laboratory used the bioassay and dissociative extraction techniques to progressively purify bone-inductive proteins from bovine bone matrix. This work led to the isolation and partial characterization of osteogenin, one of the first individual BMP proteins to be separated from the complex mixture Urist had originally described.(12)

Reddi's laboratory also established that BMPs were pleiotropic regulators acting in a concentration-dependent manner, that they bind the extracellular matrix, and that they are present at the apical ectodermal ridge during limb development — connecting the clinical biology of bone repair to the fundamental biology of skeletal morphogenesis.(12) The conceptual advance was significant: BMP was not just a repair signal. It was a morphogen — a substance that, depending on concentration and context, could instruct cells to take on fundamentally different skeletal fates.

The molecular mechanism of bone induction led to the conceptual advance in tissue engineering that morphogens bound to an insoluble extracellular matrix scaffolding act in collaboration to stimulate stem cells to form cartilage and bone.

III.  John M. Wozney

The Man Who Cloned the Signal

A Molecular Biologist Enters Bone Biology

John Wozney came to the BMP problem not through orthopedics but through molecular biology. Working at the Genetics Institute in Cambridge, Massachusetts — a biotechnology company founded in 1980 by Harvard molecular biologists Thomas Maniatis and Mark Ptashne — Wozney led the effort to take Reddi's partially purified proteins and use recombinant DNA technology to identify, clone, and produce them at scale.(11)

His approach was to use amino acid sequence information extracted from partially purified bovine bone proteins to design oligonucleotide probes, then use those probes to screen human cDNA and genomic libraries. The strategy was technically demanding but conceptually straightforward: find the gene, clone it, express the recombinant protein, and test whether it retained bone-forming activity.(5)

The 1988 Science Paper

The results appeared in Science in December 1988. Wozney, Rosen, Celeste, and colleagues reported the molecular cloning and characterization of BMP-1 through BMP-4 from human cDNA libraries.(5) BMP-1 turned out to be a metalloprotease — structurally unrelated to the others. BMP-2, BMP-3, and BMP-4, however, were novel members of the transforming growth factor-beta superfamily, confirming the classification work Reddi's laboratory had been advancing.(5)

BMP-2 and BMP-4 were 92% identical at the amino acid level — closely related enough to suggest similar functions, distinct enough to merit individual characterization. Recombinant BMP 2 and 4 retained osteoinductive activity in vivo.(5) For the first time, a defined, sequence-characterized, recombinantly produced protein could reliably instruct mesenchymal progenitor cells to form bone.

Two years later, Celeste et al. extended the family further, identifying BMP-5 through BMP-8 from sequence homology screening.(6) The BMP family, it became clear, was large — eventually reaching more than twenty members — and its biology extended well beyond bone, into kidney development, neural patterning, and embryonic morphogenesis. What Urist had called the "bone inductive principle" turned out to be a pleiotropic family of morphogens with roles throughout vertebrate development.

Novel regulators of bone formation: molecular clones and activities.

From Clone to Clinic

The cloning of BMP-2 opened the path to pharmaceutical-scale production. Through the early 1990s, Genetics Institute — later acquired by Wyeth — scaled up recombinant human BMP-2 production using Chinese hamster ovary (CHO) cell expression systems, the standard platform for complex glycoprotein biologics.(14)

The subsequent decade of preclinical work focused on three questions Reddi's cascade model had already framed: which carrier, what concentration, and at what site. The absorbable collagen sponge (ACS) emerged as the preferred delivery vehicle — capable of localizing the protein, providing a scaffold environment analogous to the matrix component Sampath and Reddi demonstrated was necessary for osteoinduction.(13,14)

In 2002, the FDA approved INFUSE Bone Graft — rhBMP-2 delivered on an ACS, used in combination with a tapered titanium fusion cage — for single-level anterior lumbar interbody fusion between L4 and S1.(13) The pivotal randomized controlled trial demonstrated a 24-month fusion rate of 94.5% with rhBMP-2 versus 88.7% with iliac crest autograft.(13) It was the first recombinant growth factor approved by the FDA for skeletal repair — the direct clinical descendant of Urist's 1965 muscle pouch experiments.

IV.  The Signal and the Market

The BMP story follows the same arc as the MSC story told in Part One. A genuine scientific advance, carefully developed over decades, became a commercial phenomenon — and the gap between what the science demonstrated and what the market promoted widened in ways that caused real harm.

rhBMP-2 was approved for one indication: single-level ALIF between L4 and S1 with a specific titanium cage, at 1.5 mg/mL, on an ACS.(13) Within a few years of approval, off-label use had expanded to posterior and transforaminal approaches, cervical fusion, pediatric cases, multi-level constructs, and open fractures — applications the original clinical trials had never examined.(14)

By 2008, serious complications had begun appearing in the published literature: ectopic bone formation, osteolysis, vertebral endplate resorption, retrograde ejaculation, soft tissue swelling severe enough to compromise the airway in anterior cervical cases. The FDA issued a public health notification on cervical use.(15) In 2013, the Yale University Open Data Access (YODA) Project completed an independent review of all published and unpublished rhBMP-2 data. Its conclusions were measured: rhBMP-2 was comparable to iliac crest autograft for fusion rates, but the original industry-sponsored trials had underreported complications, and some analysis suggested a possible increase in cancer signal that warranted continued monitoring.(15)

The complications were not random. They were, in most cases, predictable extensions of the biology. Reddi's cascade model had established that BMP is a concentration-dependent morphogen — dose too little and nothing happens; dose too much and the signal overruns its intended boundaries. Delivering a supraphysiologic quantity of rhBMP-2 on a rapidly releasing absorbable sponge in a confined anatomical space is not how the signal operates in vivo. The underlying biology remained valid, but clinical outcomes proved highly dependent on dose, carrier, anatomy and application technique.

The lesson runs parallel to the MSC overclaiming story: the signal that instructs progenitor cells to form bone is real, clinically meaningful, and still in use today. The problems arose not from the science but from the space between a narrow FDA approval and an expansive commercial deployment.

V.  What "The Signal" Means for Clinical Practice

The work of Urist, Reddi, and Wozney established four principles that remain as relevant to current osteobiologic practice as the cellular principles established in Part One.

1

Osteoinduction is a cascade, not a switch. Reddi's sequence — progenitor recruitment, proliferation, differentiation, hypertrophy, vascularization, bone formation — means that disrupting any step compromises the outcome. Patient factors that impair any stage of this process (poor vascularity, immunosuppression, smoking, metabolic disease) are not incidental to biologic performance. They are biologically fundamental.(3,12)

2

Signal and scaffold are inseparable. Sampath and Reddi's two-component finding explains why carrier selection in any biologic/carrier combination product is not a packaging decision — it is a biological one. The scaffold provides the local matrix environment that strongly influences signal retention, presentation and biological activity. Additionally, carrier properties (resorption rate, porosity, surface chemistry) directly determine biologic performance.(4)

3

Dose and site determine outcome. BMP is a concentration-dependent morphogen. The complications associated with rhBMP-2 — ectopic bone, osteolysis, soft tissue swelling — are largely dose- and site-related consequences of supraphysiologic delivery in confined anatomical spaces. The biology of the signal does not change; the risk changes with how it is delivered.(14,15)

4

The BMP family is large — and largely untapped. Twenty or more BMP family members have been identified. Clinical use of only BMP-2 and BMP-7 represents a narrow slice of an extensive biological toolkit. Next-generation BMP-based therapies — including designer chimeric BMPs with improved receptor selectivity and reduced ectopic activity — are in active development.(5,6,16)

NEXT IN THE SERIES: PART THREE — "THE CLINIC"

Part Three traces how Scott Boden, George Muschler, and Philippe Hernigou took the cells identified in Part One and the signals identified in Part Two, and built the clinical protocols and evidence base that brought osteobiologics into the operating room — from the first validated spinal fusion BMP model to the science of bone marrow aspiration technique.

Closing Thought

Bone holds its own building instructions inside its matrix — a set of molecular signals often present in parts per billion, embedded there long before any surgeon or scientist thought to look for them. Urist found them by accident and spent thirty years trying to isolate them. Reddi decoded how they work and established that they cannot function without a scaffold. Wozney gave them a sequence, a name, and a manufacturing process.

What followed was predictable in retrospect: a powerful biology was extracted from its natural context, concentrated, delivered in ways its authors had not intended, and used in patients for whom the evidence provided no guidance. Some of those patients were helped. Some were harmed. The signal itself was never the problem. The problem was the gap between understanding what the signal does and understanding the full range of what it does when you change the dose, the carrier, and the site.

The underlying biology was never invalid.  The challenge was understanding how a physiologic developmental signal behaves when delivered at supraphysiologic doses, on synthetic carriers, in anatomical environments very different from those in which it evolved.

References

1.  Urist MR. Bone: formation by autoinduction. Science. 1965;150(3698):893–899. doi:10.1126/science.150.3698.893

2.  Urist MR, Strates BS. Bone morphogenetic protein. J Dent Res. 1971;50(6):1392–1406. doi:10.1177/00220345710500060601

3.  Reddi AH, Huggins CB. Biochemical sequences in the transformation of normal fibroblasts in adolescent rats. Proc Natl Acad Sci USA. 1972;69(6):1601–1605. doi:10.1073/pnas.69.6.1601

4.  Sampath TK, Reddi AH. Dissociative extraction and reconstitution of extracellular matrix components involved in local bone differentiation. Proc Natl Acad Sci USA. 1981;78(12):7599–7603. doi:10.1073/pnas.78.12.7599

5.  Wozney JM, Rosen V, Celeste AJ, et al. Novel regulators of bone formation: molecular clones and activities. Science. 1988;242(4885):1528–1534. doi:10.1126/science.3201241

6.  Celeste AJ, Iannazzi JA, Taylor RC, et al. Identification of transforming growth factor beta family members present in bone-inductive protein purified from bovine bone. Proc Natl Acad Sci USA. 1990;87(24):9843–9847. doi:10.1073/pnas.87.24.9843

7.  Sampath TK, Reddi AH. Homology of bone-inductive proteins from human, monkey, bovine, and rat extracellular bone matrix. Proc Natl Acad Sci USA. 1983;80(21):6591–6595. doi:10.1073/pnas.80.21.6591

8.  Grgurevic L, Oppermann H, Pecina M, Vukicevic S. Marshall R. Urist and the discovery of bone morphogenetic proteins. Int Orthop. 2017;41(5):1065–1069. doi:10.1007/s00264-017-3402-9

9.  Reddi AH, Urist MR. A renaissance scientist and orthopaedic surgeon. J Bone Joint Surg Am. 2003;85-A(Suppl 3):3–7.

10.  Sampath TK, Reddi AH. Discovery of bone morphogenetic proteins — a historical perspective. Bone. 2020;140:115548. doi:10.1016/j.bone.2020.115548

11.  Katagiri T, Tsukamoto S, Nakachi Y, Kuratani M. Discovery of heterotopic bone-inducing activity in hard tissues and the TGF-β superfamily. Int J Mol Sci. 2018;19(11):3586. doi:10.3390/ijms19113586

12.  Reddi AH. Role of morphogenetic proteins in skeletal tissue engineering and regeneration. Nat Biotechnol. 1998;16(3):247–252. doi:10.1038/nbt0398-247

13.  Burkus JK, Gornet MF, Dickman CA, Zdeblick TA. Anterior lumbar interbody fusion using rhBMP-2 with tapered interbody cages. J Spinal Disord Tech. 2002;15(5):337–349. doi:10.1097/00024720-200210000-00001

14.  Malham GM, Louie PK, Brazenor GA, Mobbs RJ, Walsh WR, Sethi RK. Recombinant human bone morphogenetic protein-2 in spine surgery: recommendations for use and alternative bone substitutes — a narrative review. J Spine Surg. 2022;8(4):477–490. doi:10.21037/jss-22-23

15.  Carragee EJ, Hurwitz EL, Weiner BK. A critical review of recombinant human bone morphogenetic protein-2 trials in spinal surgery: emerging safety concerns and lessons learned. Spine J. 2011;11(6):471–491. doi:10.1016/j.spinee.2011.04.023

16.  Wozney JM. Overview of bone morphogenetic proteins. Spine. 2002;27(16 Suppl 1):S2–8. doi:10.1097/00007632-200208151-00002