Since the approval of the random amino acid copolymer YEAK (Copaxone®) for the treatment of RR-MS, YFAK, a second generation copolymer with enhanced efficacy in mouse models of EAE has been described. Understanding the mechanism of action of the copolymers has engaged numerous laboratories (reviewed by 
). Here, YFAK and YEAK are shown to have dramatically different effects on the immune system in mice, especially with respect to effects on the innate immune response that plays an important role in efficacy. The major molecular differences between YFAK and YEAK are highlighted in . YFAK is a solid phase synthesized acetylated 52mer random copolymer (relative ratio: Tyrosine 1.0, Phenylalanine 1.2, Alanine 23.5, and Lysine 6.0) that has a strong net positive charge. YEAK, on the other hand, is a solution phase synthesized non-acetylated 20–200mer random copolymer (relative ratio: Tyrosine 3.5, Glutamic Acid 1.5, Alanine 4.5, and Lysine 3.6) that has a slightly net positive charge. YFAK is significantly more efficacious than YEAK in the prophylaxis of the induction of EAE or in the treatment of established EAE in mice (
, and further documented in more detail in Figure S3
Molecular Properties of YFAK and YEAK.
The serum levels reached in mice after s.c. administration (measured using a newly developed antibody-based method) of YFAK and its duration were much larger than those for YEAK (). In spite of its peptidic nature, YFAK (and sub-fragments ≥30-mers) were detectable in mouse serum for several hours. YFAK may be binding to a plasma component that prevents its rapid degradation by peptidases, possibly a function of its strong net positive charge when compared to YEAK. Acetylation of YFAK on the N-terminus also may help protect it from peptidase degradation. A significant linear correlation was observed between the mean serum concentration of YFAK or YEAK (Cmax) and the administered s.c. dose.
The detection of YFAK and YEAK in the serum coincided with the appearance in plasma of CCL22 and CXCL13 within minutes of copolymer administration (). These chemokines are known to be secreted by regulatory M2a and M2c macrophages, respectively 
, also known as alternatively activated macrophages 
. YFAK induced a significantly higher Cmax
for CCL22 and CXCL13 than YEAK. Production of CCL22 and CXCL13 induced by both copolymers was independent of MHC class II as shown by their secretion in the MHC Class II−/− mice (). A second innate immune receptor for these peptide copolymers, in addition to MHC Class II, must be functional. Systemic exposure to copolymers may be important to insure a broad effect on the innate immune system. The enhanced efficacy of YFAK in mice may be related to its intrinsic action on myeloid cells and the enhanced time of serum availability. Although the effects of YFAK and YEAK on plasma CCL22 and CXCL13 during induction of EAE and its therapy was not measured in these experiments, the analysis of serum CCL22 during administration of YFAK to patients with secondary progressive multiple sclerosis in a Phase Ib clinical trial was carried out and the appropriate elevation was observed 
Myeloid progenitors differentiated to macrophages in vitro in the presence of IL-3, as well as the macrophage cell line RAW264.7, were also shown to secrete CCL22 in response to YFAK and to YEAK at high concentrations ( and ). Moreover, splenic T cells induced by YFAK also secrete IL-3 (), as well as the M2 macrophage differentiating cytokines IL 4 and IL-13. IL-3 is a cytokine that stimulates development of multiple lineages of hematopoietic cells including myeloid macrophages and neutrophils, but excluding lymphoid cells. These data lead to the hypothesis summarized in that an important role of regulatory macrophages stimulated by amino acid copolymers is to secrete the chemokine CCL22 that attracts T cells, leading to increased numbers of IL-10-secreting T cells. Whether this phenomenon results from expansion of a small, preexisting pool of these cells or whether the regulatory macrophages induce naïve T cells to differentiate into IL-10-secreting T cells remains a subject for further study. In turn the T cells secrete IL-13, IL-4, and IL-3 that stimulate the proliferation and differentiation of M2 macrophages.
Hypothesis: Role of M2 macrophages in generation of IL-10 secreting T cells.
The induction of these Tregs by YEAK 
and YFAK 
are both well documented, and the latter have been extensively characterized. YFAK programs bone marrow progenitors to induce a population of CD11b+
macrophages, called myeloid suppressors, that can modulate an autoimmune response (Figures S1
). This phenotype for myeloid suppressor cells has been reported previously 
Some differences in cytokine/chemokine production were observed between cells cultured in the presence of YFAK or YEAK (). YFAK significantly increased supernatant concentrations of CCL22 (MDC) and of CXCL13 (BCA-1). CCL22 and CXCL13 are chemokines produced by M2 macrophages which are important in the homeostasis of lymphocyte trafficking and are a chemoattractant for T cells and B cells, respectively 
. CCL9 (MIP-1γ), an important chemokine that attracts CD11b+
DC and myeloid cells 
, was also increased as a result of YFAK treatment (Figure S4
). Thus, CCL9 may play a role in the attraction/induction of myeloid suppressor cells which have been shown to be important in decreasing disease severity 
Some data suggest that CCL22 and CXCL13 may be produced by different cell populations. The serum level of CXCL13 induced in vivo by YFAK was greater than that of CCL22 ( and ). However, in vitro the level of CCL22 induced from myeloid progenitors was 50–100 times that of CXCL13, and YEAK did not induce CXCL13 from these cells in a detectable amount (). Finally, although both YFAK and YEAK induced CCL22 secretion from RAW 264.7 murine macrophages (), neither copolymer induced CXCL13 from these cells.
In addition, YFAK significantly decreased the secretion of the chemokines CXCL1 and CXCL2 and of pro-inflammatory cytokines TNF-α, IL-6, and IL-12p70. The pro-inflammatory properties associated with TNF-α play a major role in autoimmune diseases and interference with TNF- α production is a major treatment modality 
. Etanercept (Enbrel®), infliximab (Remicade®), and adalimumab (Humira®) are TNF inhibitors that are used in animal disease models of MS 
and have been approved for treatment of patients with rheumatoid arthritis, inflammatory bowel disease, and psoriasis 
. IL-12p70 has also been shown to drive a TH
1 immune response 
and decreasing IL-12p70 production can modulate the immune response towards a TH
2 immune response 
. Since modulation of pro-inflammatory cytokines is associated with current MS treatment 
, this effect of YFAK could directly relate to its efficacy in EAE. By contrast, YEAK had little or no effect on secretion of these chemokines and cytokines in various assays.
The relationship between dose of YFAK and efficacy in EAE is of considerable interest, i.e., 0.75 mg/kg is inefficient at generating efficacy, 2.5 mg/kg gives significant efficacy, and at 7.5 mg/kg efficacy is lost. As the dose of YFAK increased, so did activation of the innate immune response (). However, in vitro CCL22 secretion reached a peak of 6 µM and then declined. YEAK, which is ineffective at low concentration, peaked at 12 µM. If higher concentrations of YEAK had been used, a decline in CCL22 secretion may also have been seen. A decline in CCL22 secretion at high YFAK concentrations was also seen in an additional in vitro
assay (N.K., unpublished observation), and was paralleled in Phase I clinical trials in secondary progressive MS patients by increased serum concentrations of CCL22, IL-3, and IL-13 at low dosages followed by decreases at higher dosages 
. However, as the dose of either copolymer increased, a progressive loss in IL-3 production from T cells occurred (). T cell activity alone is not solely responsible for efficacy because YFAK and YEAK would both have similar efficacy and the lower the dose the higher the T cell activity, i.e., 0.25 mg/kg had more IL-3 production than 2.5 mg/kg. However, 2.5 mg/kg is more efficacious for therapy. The innate immune response is also not solely responsible for efficacy. Together these data imply that both the innate (macrophages) and adaptive (T cells) immune responses are responsible for efficacy.
Interaction between cells of the innate and adaptive immune system is quite important in the mechanism. Th17 cells are believed to play a major role in autoimmune pathologies and multiple sclerosis in particular. YEAK has been shown to dampen differentiation of Th17 cells through altered production of IL-6 by monocytes 
. YFAK reduces IL-6 secretion more severely () and may effectively control the expansion and differentiation of Th17 cells. IL-13, a cytokine produced in large quantity by T cells activated with both YEAK and YFAK, has a major effect on the differentiation of monocytes 
. Moreover, production of IL-13 by these T cells could directly or indirectly (through antigen-presenting cells) affect the expansion of Th17 cells, since it is known that the IL-13 receptor is expressed by Th17 cells and that IL-13 attenuates Th17 cytokine production 
. Thus, the optimal dosage requirement for YFAK is one that strongly stimulates the innate response, yet is low enough to not abrogate an adaptive immune response.
In summary, chemokine release from myeloid cells, which occurs rapidly on administration of amino acid copolymers in vivo, has significant implications for the systemic effect of these therapeutic agents on both innate and adaptive immunity. The differential effects of YFAK and YEAK on the innate immune system may explain the enhanced efficacy of YFAK in the treatment of EAE and is also likely to lead to better clinical efficacy.