A zinc-finger protein (Zn is silver) recognizes and binds exact DNA sequence. Sangamo engineers zinc-fingers to edit and correct genes carrying disease-causing mutations.

On October 3, 2011 Sangamo Biosciences announced it was halting its lead program SB-509 program in diabetic neuropathy after the zinc-finger protein (ZFP) therapy failed to meet its end points. Although the failure represents a considerable setback for the diabetic neuropathy program, Sangamo remains buoyant about ZFPs as a therapeutic platform, arguing that challenges associated with SB-509’s ambitious target indication conspired to thwart the drug.

Sangamo has built its name as the pioneer of engineered ZFPs as therapeutics. ZFPs are DNA-binding proteins that attach to unique sequence motifs and participate in gene regulation. Each ‘finger’ recognizes a particular three-base set (GNN and ANN triplets) and by choosing the proper number and arrangement of motifs, one can construct specialized ZFPs that promote targeted regulation. What’s more, considerable interest focuses around the potential of ZFPs, when joined through a linker to an endonuclease like Fok1, in editing or correcting endogenous genes containing disease-causing mutations.

The ZFP gene activator employed in SB-509 was first described in. SB-509 is a plasmid encoding three ZFPs that target a GGGGGTGAC site in the vascular endothelial growth factor A (VEGFA) gene. The ZFPs are linked to a p65 transcriptional activator that is constitutively expressed by a cytomegalovirus early promoter/enhancer and switches on VEGFA after binding. The ZFPs are injected intramuscularly in a formulation containing Poloxamer 188, sodium chloride and tris-HCl.

According to Alastair Mackay, an analyst at GARP Research & Securities, in Baltimore, many investors were already tentative about the ZFP technology. “The outside community was split,” he says, “with bears saying for some time that this latest trial would not succeed, while bulls were saying that the trial had good prospects for success.” To be sure, peripheral diabetic neuropathy is an ambitious target. This common complication of diabetes is actually a heterogeneous syndrome of pathologies, in which patients gradually accumulate damage to the microvasculature in their extremities. This leads to reduced neuronal input and poorer signal conduction in affected tissues, causing pain and/or increasing loss of sensation and leaving diabetic neuropathy sufferers vulnerable to serious injury and infection. The only available treatments are palliative, a combination of analgesics and antidepressants, and attempts to develop efficacious therapeutics have consistently failed.

Although VEGFA modulation could theoretically address a host of disease conditions, Sangamo CEO Edward Lanphier cites several reasons why his company saw diabetic neuropathy as the right indication. “VEGFA is first and foremost a potent neurotrophic factor,” he says. “Peripheral diabetic neuropathy involves both nerve loss as well as loss of microvascular structure, and the vascular and neurotropic functions made VEGFA ideally suited for treating this disease.” According to Lanphier, rodent studies indicated that SB-509 could promote revascularization and repair, potentially accelerating the time required to see symptom improvement. Importantly, SB-509 seemed to promote production of all of the various VEGFA isoforms, increasing the chances for the ZFP to elicit a full spectrum of VEGFA’s effects.

In 2006, the Juvenile Diabetes Research Foundation (JDRF) partnered with Sangamo, providing financial support for their ‘601’ phase 2 trial, which enrolled 110 patients with mild to moderate peripheral diabetic neuropathy. The therapy proved safe, with only minimal adverse effects, although efficacy results were a mixed bag and patients in the treatment arm as a whole failed to achieve significant benefit relative to the placebo group.

It didn’t help that in the vast majority of people, diabetic neuropathy can be forestalled with glucose control, explains Helen Nickerson, a member of the JDRF scientific staff for complications therapies. “The placebo-group patients are very slow to decline,” she says, which adds a layer of difficulty to trial designs.

A subset of patients, however—those with more severe disease—did experience some benefit in an array of neurological indicators. Ted Tenthoff, a managing director and senior research analyst at New York–based securities firm Piper Jaffray, and other observers were unconvinced. Tenthoff viewed the effort to carve out a new primary end point from the unsuccessful phase 2 results as ‘data mining’, and his firm adopted a more pessimistic view of the company’s prospects. “We felt like the risk around SB-509 clinically was too great for the valuation,” he says.

Despite such concerns, Sangamo’s development team opted to test SB-509 in a more carefully chosen patient cohort. “Looking at the data from the first study, it was very clear—and highly statistically significant based on a retrospective analysis—that there was a group of patients with moderate severity disease who showed the greatest improvement versus placebo,” says Lanphier. JDRF renewed their support, providing the funding for the ‘901’ phase 2b trial, which recruited 170 patients with moderate or severe peripheral diabetic neuropathy and incorporated a host of outcome measures, such as the neuropathy impairment score–lower limb, broadly used in clinical settings.

With the results of this latest ‘901’ trial unequivocally disappointing, Sangamo has ceased further development of SB-509. To Lanphier’s thinking, the primary culprit, as in previous trials, was that placebo patients simply remained too healthy to observe meaningful relative benefits from treatment. “With the new oral medications that help stabilize blood sugar levels, there’s better care for patients in general with type 2 diabetes,” he says, “but also there’s simply better care for patients who are in a clinical trial.” In a press release after the conclusion of the 901 trial, the company noted that the treatment arm achieved neurological improvements equivalent to responders in the 601 trial but that these were masked by the positive response of the placebo arm to routine medical treatment.

Others point out potential technical complications as well, criticizing the decision to deliver SB-509 through a series of intramuscular injections of naked plasmid DNA. “Delivery via plasmid injection hasn’t lived up to its promise even in vaccines, which require more limited expression than this therapeutic application,” says Carlos Barbas III, a pioneer in the field of zinc-finger engineering at the Scripps Research Institute in La Jolla, California. “I don’t think plasmid injection gave SB-509 a fighting chance.” Seppo Ylä-Herttuala, whose group at the University of Kuopio in Finland has worked extensively with VEGF-oriented gene therapy, also expressed concern about the chosen method, but acknowledges that choices are limited. “We have two or three other options at the moment, but they all have the same issues related to delivery,” he says. Lanphier, however, defends the decision, pointing out that plasmid injection was well suited to Sangamo’s primary goal of achieving short bursts of transient expression by means of chronic dosing.

Some also question whether the effects of SB-509 were undercut by insufficient specificity. With only three zinc fingers, the construct targets a nine-base target sequence—far from unique in the human genome. Cell culture and rodent studies showed limited off-target effects, but Ylä-Herttuala is skeptical about those results. “In our hands, with roughly equivalent but non-VEGF zinc-finger constructs, we’ve seen as many as a hundred genes going up or down,” he says.

Although none of these factors alone may explain the failure of SB-509, collectively they appear to have spelled its doom. “They took on a very tough disease with a very tough approach,” says Tenthoff.

Sangamo is now moving ahead with its SB-728-T program, which targets chemokine CC-motif receptor 5 (CCR5), a co-receptor for HIV. SB-728-T comprises two endonuclease domains linked to a pair of ZFPs, each containing four zinc-finger motifs (recognizing a total of 24 base pairs in the CCR5 gene). Concerted binding of the DNA induces dimerization of the two endonuclease domains, resulting in a double-stranded break of the CCR5 sequence. This in turn induces cellular DNA repair pathways, most notably the mutagenic nonhomologous end-joining pathway, leading to efficient disruption of the CCR5 gene. The approach used for HIV treatment involves the isolation of CD4⁺ T cells from HIV-positive patients and their treatment ex vivo with SB-728. Treated cells are then reinfused in the patient, with the hope that virus-resistant T cells will expand to become the patient’s main population.

The company presented phase 1 data this past September at the Interscience Conference on Antimicrobial Agents and Chemotherapy in Chicago, showing potential patient benefit, particularly for one individual who was already heterozygous for the naturally occurring CCR5Δ32 mutation. The company is embarking on several additional trials, including one that will specifically assess the efficacy of SB-728-T in a larger cohort of heterozygous patients.

In the meantime, Sangamo is likely to remain out of favor with analysts. “The problem that Sangamo faces is not that their results are bad—they’re good—but they’re phase 1 results,” says Mackay. In the same vein, a spate of high-profile research articles demonstrating the potential of zinc-finger nucleases to correct genomic defects in monogenic diseases like hemophilia show considerable promise. But the US Food and Drug Administration’s cautious approach to gene therapy trials means that even under ideal circumstances, the company must run a long and expensive gauntlet to get these experimental treatments off the ground.

At an investor teleconference at the end of its third quarter in 2011, Sangamo reported cash reserves of at least $85 million, which they project as being sufficient to capitalize the next few years’ R&D efforts. Sangamo also maintains additional revenue streams through commercial partnerships, including a licensing agreement that enables St. Louis–based Sigma-Aldrich to sell ZFP-related research tools and an arrangement with Dow AgroSciences of Indianapolis, which has enabled that company to use Sangamo technology to generate genetically modified (not necessarily transgenic) crops. Royalties and milestone payments from such arrangements have yielded modest but steady revenues—projected at $10–12 million for 2011—and Sangamo intends to further expand on these licensing agreements in the future, although human medicine remains the priority. “Within these walls, Sangamo will stay almost exclusively focused on therapeutic applications,” says Lanphier.

As the only company actively developing ZFP therapeutics, Sangamo retains a remarkable head start. “In terms of a commercial presence, they hold a complete patent portfolio and have for many years,” says Mackay. Paris-based Cellectis is working with other classes of engineered proteins, such as meganucleases and transcription activator–like effector nucleases (TALENs) for targeted genome editing in cell biology and agriculture applications, but does not appear to be heavily focused on clinical efforts at present. Barbas also points out that TALENs are still relatively new, and questions remain as to whether they can offer the same strong safety profile that has been demonstrated for ZFPs in humans.

In the meantime, the jury remains out as to whether Sangamo can find the right indication, right molecule, right clinical trial design and most compelling data package for regulators that will lead to the registration of a completely new kind of therapeutic modality before their cash reserves run dry. “We’re certainly not where we would be if the data had been positive,” says Lanphier, “but you go forward with what you have and you do it for the right reasons, and when your expectations aren’t fulfilled, you just have to start up the hill again.”

Source: Nature Biotechnology