Measurement of absolute abundance of crystallins in human and αA N101D transgenic mouse lenses using 15N-labeled crystallin standards

Stable isotope labeled standards of all major human lens crystallins were created to measure the abundance of lens endogenous crystallins from birth to adulthood. All major human crystallins (αA, αB, βA2, βA3/A1, βA4, βB1, βB2, βB3, γA, γB, γC, γD, γS) were cloned with N-terminal 6 x His tagged SUMO for ease of purification and the ability to generate natural N-termini by SUMO protease cleavage when producing crystallins for structure/function studies. They were then expressed in ¹⁵N-enriched media, quantified by mass spectrometry, and mixed in proportions found in young human lens to act as an artificial lens standard. The absolute quantification method was tested using soluble protein from 5-day, 23-day, 18-month, and 18-year-old human lenses spiked with the ¹⁵N artificial lens standard. Proteins were trypsinized, relative ratios of light and heavy labeled peptides determined using high-resolution precursor and data independent MS2 scans, and data analysis performed using Skyline software. Crystallin abundances were measured in both human donor lenses and in transgenic mouse αA N101D cataract lenses. Technical replicates of human crystallin abundance measurements were performed with average coefficients of variation of approximately 2% across all 13 crystallins. αA crystallin comprised 27% of the soluble protein of 5-day-old lens and decreased to 16% by 18-years of age. Over this time period αB increased from 6% to 9% and the αA/αB ratio decreased from 4.5/1 to 2/1. γS-crystallin also increased nearly 2-fold from 7% to 12%, becoming the 3rd most abundant protein in adult lens, while βB1 increased from 14% to 20%, becoming the most abundant crystallin of adult lens. Minor crystallins βA2, βB3, and γA comprised only about 1% each of the newborn lens soluble protein, and their abundance dropped precipitously by adulthood. While 9 of the SUMO tagged crystallins were useful for purification of crystallins for structural studies, γA, γB, γC, and γD were resistant to cleavage by SUMO protease. The abundance of WT and N101D human αA in transgenic mouse lenses was approximately 40-fold lower than endogenous mouse αA, but the deamidation mimic human αA N101D was less soluble than human WT αA. The high content of αA and the transient abundance of βA2, βB3, and γA in young lens suggest these crystallins play a role in early lens development and growth. βB1 becoming the most abundant crystallin may result from its role in promoting higher order β-crystallin oligomerization in mature lens. The full set of human crystallin expression vectors in the Addgene repository should be a useful resource for future crystallin studies. ¹⁵N labeling of these crystallins will be useful to accurately quantify crystallins in lens anatomic regions, as well as measure the composition of insoluble light scattering crystallin aggregates. The standards will also be useful to measure the abundance of crystallins expressed in transgenic animal models.