Indoleamine 2,3-Dioxygenase Overexpression Causes Kynurenine-modification of Proteins, Fiber Cell Apoptosis and Cataract Formation in the Mouse Lens

Maneesh Mailankot; Magdalena M. Staniszewska; Heather Butler; Moonkyung H. Caprara; Scott Howell; Benlian Wang; Catherine Doller; Lixing W. Reneker; Ram H. Nagaraj


Lab Invest. 2009;89(5):498-512. 

In This Article


The purpose of this study was to investigate the function of KYNs in lens protein modifications. We found that overexpression of IDO in the lens results in enhanced KYN formation. Thus, enough L-tryptophan must have been available to support the increased KYN formation. This is further supported by the finding that in organ-cultured homTg lenses, adding L-tryptophan enhanced KYN formation. We did not determine the amount of NFK formed, but because we detected higher levels of KYN in homTg animals as compared to hemTg or Wt animals, it suggests that kynurenine formamidase was not rate limiting. However, it is also possible that NFK was nonenzymatically converted to KYN. Second, the KYN produced from IDO reacted with proteins to form adducts that were detected by an mAb directed against KYN modification. We also found that KYN modification occurred in the cytosol and nuclei of fiber cells, and cells containing KYN modifications underwent apoptosis. Finally, IDO overexpression caused severe morphological changes in the lens. All our studies were carried out with 3-month-old lenses, the homTg lenses had dense cataracts at this time. The question arose as to whether cataract formation was related to KYN formation. Our data on lenses from 1-month-old homTg mice showed that the KYN content and KYN modification of proteins along with fiber cell apoptosis and GSH loss were already increased at this age. This suggests that further increase in these changes could have resulted in cataract formation in 3-month-old lenses (Figure 5).

KYNs must be deaminated to produce adducts in proteins. GSH is believed to inhibit such adduction by binding to KYN. Mammalian lenses, including those in human, have high levels of GSH.[32] However, in aging and cataractous lenses, GSH levels are reduced,[33,34] which could lead to increased KYN modification of proteins. Our data showing that GSH levels are significantly reduced in homTg lenses further support this view. It is also possible that reactive oxygen species produced from KYNs[35] could have oxidized GSH and promoted reaction between KYN and proteins. It is noteworthy that in ground-squirrel lenses, even though KYN levels are relatively high, the lenses are clear, possibly because the GSH level is also extremely high.[36]

It is possible that the severe morphological changes observed with IDO overexpression were due to overexpression of the protein alone rather than its associated increased activity. However, our finding that the glyoxalase-I-overexpression mice, which have the same transgene construct but normal morphology, rules out such an effect. Therefore, it is likely that enhanced IDO expression and activity led to the morphological defects seen in the IDO transgenic lenses.

The morphological changes were highlighted by improper terminal differentiation of fiber cells (ie, nucleated). During normal lens development, epithelial cells stop dividing at the lens equatorial region and differentiate into elongated fiber cells, during which time they lose their nuclei and other organelles.[37] The poor differentiation we observed suggests a serious abnormality in organelle degradation. Although the precise mechanisms for organelle degradation are not known, it is believed that a gradient formed by diffusible substances, in particular intracellular oxygen, is one cause.[38] It is possible that in the homTg lenses, such a gradient was altered and resulted in the loss of cues necessary for organelle degradation.

The absence of differentiation may also be due to an effect of KYNs on cell signaling, transcription, and growth factors. Fibroblast growth factor, IGF-1, and bone morphogenetic protein have been implicated in stimulation of fiber cell differentiation.[39,40,41,42] In addition, it has been demonstrated that the PI3K pathway regulates initiation of lens cell differentiation by signaling a reorganization of the actin cytoskeleton from stress fibers to cortical fibers.[43] Furthermore, targeted mutations in the genes encoding c-Maf, Sox1, and Prox1 have all resulted in specific defects in fiber cell differentiation,[44,45,46] implicating these transcription factors in epithelial cell differentiation. KYNs may have affected any one of these factors, thereby affecting fiber cell differentiation. Such effects of KYN may be due to chemical modifications of proteins or through formation of reactive oxygen species. The mass spectrometric analysis of proteins from homTg lenses suggested that major proteins of the lens (α- and ß-crystallins) are modified by KYN. For that analysis, we selected only the major band from the SDS-PAGE of the immunoprecipitated sample (Figure 11). Other minor protein bands that were not analyzed might contain proteins that are involved in fiber cell differentiation. Further work is needed to verify this possibility.

Several previous studies have shown that extracellular KYNs can induce apoptosis,[47,48,49] but none has demonstrated that intracellular KYN formation can induce apoptosis. Our study strongly implicates intracellular KYNs, KYN modifications, or both in apoptosis. Although we did not determine whether nuclear protein or DNA modification was required, such an event is likely. The idea that KYNs induce fiber cell apoptosis is supported by a recent study showing that kidney tubular epithelial cells that overexpress IDO undergo apoptosis,[15] possibly through formation of KYNs. Another recent study showed that fiber cells undergo apoptosis in αA/αB-crystallin knockout mice, suggesting a function for α-crystallins in preventing fiber cell apoptosis.[50] Whether α-crystallin's potential ability to inhibit apoptosis was compromised in homTg animal lenses is uncertain, but KYN modification can cause the loss of chaperone function,[30,51] which is likely related to α-crystallin's anti-apoptotic activity. It is probable that KYNs affected FGF-R signaling, as it is required for early and late fiber cell differentiation and survival of terminally differentiated fiber cells.[52]

Only the homTg mice exhibited reduced eye size, which is common in transgenic mouse models that lack growth factors or proteins involved in cell-cycle regulation.[53,54] Inactivation of Rho GTPase by selective expression of the C3-exoenzyme[55] or expression of diphtheria toxin with regulatory sequences associated with γ2-crystallin gene [56] also results in a smaller lens and microphthalmia in mice. It is possible that diphtheria toxin and KYNs exert similar cytotoxic effects that lead to lens developmental defects and smaller lenses.

In the lens, IDO is present primarily in epithelial cells. In the normal transparent lens, IDO activity may be kept to a minimum so that its activity is sufficient to produce the low quantity of KYN needed to filter out UV light. Such regulation of IDO activity might occur at posttranslational level. IDO activity has been reported to be inhibited by NO and H2O2.[57,58] Both modifications are reversible and therefore could be involved in regulating IDO activity. During anterior segment infection, interferon-γ increases in the aqueous humor. If IDO activity goes unchecked due to stimulation by interferon-γ, then the enhanced IDO activity in the lens may impart considerable damage. In fact, elevated levels of aqueous humor interferon-γ concentrations are associated with cataract formation,[59] and ectopic expression of interferon-γ disrupts fiber cell differentiation and leads to cataract formation in mice.[60,61] It will be of interest to determine whether IDO activity is elevated and consequently KYN production enhanced when interferon-γ is elevated in the lens.

In summary, our study provides the first evidence that KYNs formed within the lens are cytotoxic and bring about changes that are similar to those that occur during cataract formation. This animal model may also be useful for understanding pathologies of other diseases, such as Alzheimer's and Parkinson's diseases, where KYNs appear to be involved.

Supplementary Information accompanies the paper on the Laboratory Investigation website (