**4.1 Methyl esterification of bilins**

*Autism Spectrum Disorders - Advances at the End of the Second Decade of the 21st Century*

urine specimens between ASD and controls.

begun using LC-MS/MS to screen for GEs, and other potential opiate peptides, in urine. Using specimens from urines collected and reported in [38], we observed by LC-MS/MS a significant difference in the abundance of a species with *m/z* 595 in the

As shown in **Figure 2**, the LC chromatograms of (A) control and (B) ASD urines show a remarkable difference in abundance for this species which elutes at ~12.9 min; typical depletion in ASD urines was 67% vs. the controls (in **Figure 2B** depletion is a factor of 68). The MS/MS of this species, codenamed "595A" at the time, is shown in **Figure 2C**. Clearly, the fragmentation pattern of 595A does not correspond to that of GE-B5 shown in **Figure 1** and is dominated by the formation of two product ions at *m/z* 470 and *m/z* 345, respectively. While we did not know the identity of 595A initially, intuition about fragmentation behavior of small molecules led us to suspect a particularly stable molecule, perhaps of the porphyrin family. The emergence of the Human Metabolome Database (HMDB) in 2007 [40] allowed us to search for possible metabolites, and we obtained one hit that satisfied the *m/z* of 595A and porphyrin metabolic pathways—l-stercobilin, C33H46N4O6, a metabolite found in mammalian waste products. Subsequently, we purchased stercobilin hydrochloride from Frontier Scientific and performed nanoelectrospray ionization (nanoESI) and MS/MS on stercobilin hydrochloride and obtained the same fragmentation pattern [41], and we have established this identity in a number of publications [42–45]. We had thus identified 595A as l-stercobilin as a potential biomarker of ASD in human urine. However, methods of performing reliable

*Chromatogram of (A) control urine extract and (B) ASD urine extract. (C) MS/MS of the peak at 12.9 min* 

*in the ASD urine extract, with a precursor ion mass of m/z 595.*

**54**

**Figure 2.**

In order to test whether stercobilin (or any other tetrapyrrole bilin) might be useful as a potential ASD biomarker, it would be necessary to have a reliable standard for quantitative purposes. Our first approach to develop a standard for the bilins was to perform methyl esterification [42]. To achieve methyl esterification of bilins, methanolic HCl was first prepared by the combination of 160 μL of acetyl chloride with 1 mL of anhydrous methanol dropwise *slowly*. Next, 10 μg of the bilin was dissolved in 500 μL of the methanolic HCl and reacted at room temperature for 2 h. This led to efficient (>90%), but incomplete, derivatization of stercobilin at its carboxyl groups to the dimethylester; MS/MS using collision-induced dissociation (CID) with argon gas confirmed that derivatization occurs at each carboxyl group, and although most of the bilins were converted to the dimethylester form, a small amount of the monomethyl ester was also produced. Furthermore, the dimethylesters were found to have high solubility in 50/50 methanol/water; this was found to be problematic for longterm stability, as the level of esterification would decrease over storage time at 4°C.
