Shedding light on melanins within in situ human eye melanocytes using 2-photon microscopy profiling techniques

Choroidal melanocytes (HCMs) are melanin-producing cells in the vascular uvea of the human eye (iris, ciliary body and choroid). These cranial neural crest-derived cells migrate to populate a mesodermal microenvironment, and display cellular functions and extracellular interactions that are biologically distinct to skin melanocytes. HCMs (and melanins) are important in normal human eye physiology with roles including photoprotection, regulation of oxidative damage and immune responses. To extend knowledge of cytoplasmic melanins and melanosomes in label-free HCMs, a non-invasive ‘fit-free’ approach, combining 2-photon excitation fluorescence lifetimes and emission spectral imaging with phasor plot segmentation was applied. Intracellular melanin-mapped FLIM phasors showed a linear distribution indicating that HCM melanins are a ratio of two fluorophores, eumelanin and pheomelanin. A quantitative histogram of HCM melanins was generated by identifying the image pixel fraction contributed by phasor clusters mapped to varying eumelanin/pheomelanin ratio. Eumelanin-enriched dark HCM regions mapped to phasors with shorter lifetimes and longer spectral emission (580–625 nm) and pheomelanin-enriched lighter pigmented HCM regions mapped to phasors with longer lifetimes and shorter spectral emission (550–585 nm). Overall, we demonstrated that these methods can identify and quantitatively profile the heterogeneous eumelanins/pheomelanins within in situ HCMs, and visualize melanosome spatial distributions, not previously reported for these cells.

Supplementary Note S1 2PM of HCMs and surrounding human choroidal tissue in situ. Label-free and fixed HCMs in culture and in situ (choroidal flatmounts and paraffin-embedded sections) were imaged by brightfield (Supplementary Figure S1a, S1c, S1e) and 2PM (Supplementary Figure S1b, S1d, S1f). 2-photon excitation was performed at 780 nm, the optimal intracellular melanin excitation (Supplementary Note S2). 2PM images were collected in two channels (500 -550 nm and 575 -610 nm, with red and green lookup tables applied respectively) that when overlaid, showed colocalizing pixels colored yellow.

Supplementary Note S2
Optimal 2PM laser excitation wavelength for intracellular melanins. It was initially necessary to identify the optimal 2-photon pulsed laser wavelength to excite the cytoplasmic melanin mixture within label-free and fixed heterogenous pigmented HCMs because melanin is a low quantum yield fluorophore 2,3,4,5 . To ensure that we collected fluorescence signal mostly from intracellular melanins, there should be limited excitation from other endogenous fluorophores present in HCMs and the choroidal microenvironment.
NADH and NADPH, ubiquitous source of autofluorescence in all cells, are typically excited at 740 nm for optimal excitation 6,7 . In the choroidal ECM, collagens are optimally excited at 730 nm 8,9 and elastins at 750 nm 10 .
This was identified by performing a series of 2-photon lambda (ʎ) emission spectral scans and FLIM measurements between 700 and 900 nm. The results from the emission spectral scans and derived phasor plots showed two excitation wavelengths (λex) that produced the highest mean peak emission intensity (0.44 a.u.): 780nm (Ex780nm) and 830nm (Ex830nm) (Supplementary Figure S2a). Intracellular melanin excitation at 780nm provided a more consistent emission intensity peak based on the smaller standard deviation for the measured emission intensity data (0.4SD780nm vs. 0.56SD830nm).
2PM FLIM phasor distribution derived from the HCM-localized intracellular melanin mixture was also examined between 700 and 900 nm (Supplementary Figure S2b). The derived phasors were consistently linear distributed when excited between 740 and 900 nm. When the FLIM phasor plots associated with Ex780nm and Ex830nm were examined closely, intracellular melanin excitation at Ex780nm was more comprehensive than Ex830nm at deriving a phasor distribution mapping to all regions of the heterogeneous pigmented HCMs, particularly phasors associated with the darker pigmented HCMs. Based on these findings and the more consistent fluorescence emission peak, we elected to excite the intracellular melanin mixture within HCMs at 780 nm in the current study.    Figure 2b). The peak wavelength of the spectral phasor cluster's center of mass mapped to dark eumelanin-enriched melanocytes was identified at around 635 nm; for light pigmented, pheomelanin-enriched melanocytes, this was identified at around 589 nm. The shift in peak emission identified between dark eumelanin-enriched and light pheomelanin-enriched melanocytes was thus around 45 nm, a wider spread than expected if this was a melanin concentration dependent shift. melanocytes was around 45 nm, a wider spread than the expected melanin concentration dependent shift of around 10 nm. The 2PM spectral signal from HCM cells was improved by increasing acquisition bin size from 8.9 nm to 26.8 nm. However, this results in a three-fold decrease in spectral resolution and therefore, would produce poorer segmentation of closely emitting species. Therefore, the bin size for the 2PM spectral acquisition was maintained at 8.9 nm. G = X coordinate of phasor transform ('real' unitless phasor component), S = Y coordinate of phasor transform ('imaginary' unitless phasor component).   (Supplementary Figure 3).
These were excited at 780 nm and used for subsequent phasor segmentation analysis of HCM-localized melanin and choroid tissue. The merged phasors from the melanin sources (Supplementary Figure 3a, 3b, 3c, 3d)

Supplementary Note 5 Average lifetimes of melanin controls comparable with average lifetimes measured from
HCMs. We fitted the fluorescence decay measured from the eumelanin-enriched dark brown hair, dark pigmented HCM, the pheomelanin-enriched red hair and light pigmented HCM using a two exponentials model (Supplementary using non-descanned and descanned detection. Both images were contrast enhanced to 100 greyscale value. The non-descanned image showed a higher signal to noise compared to the descanned image (Supplementary Figure 5). Poorer signal to noise ratio would be the limiting factor to how many fluorophore species can be reliably unmixed using spectral phasor analysis 12 .
Supplementary Figure 5: Although non-descanned and descanned measurements are not directly comparable, descanned detection suffered from poorer signal to noise ratio even when data were acquired from the same field of view and acquisition parameters.

Supplementary Note 6 Examination of phasor plots from fixed HCM and HEK293 cells excited at 740 nm.
Cytoplasmic NADH within the melanin-free HEK293 cells was optimally excited at 740 nm, and 2PM FLIM data acquisition and phasor analysis were also performed at this wavelength on fixed HCMs. Segmented phasor clusters mapping to the light pigmented melanin mixture and cytoplasmic NADH were distinct and positioned close to each other (Supplementary

Supplementary Note 9
Other non-destructive and non-invasive approaches used to study the intracellular melanin content in skin and hair. Other complementary non-destructive and non-invasive approaches have been used to study the intracellular melanin content in skin and hair 13,14,15 . Eumelanin and pheomelanin have distinct photoionization thresholds and vastly different transient absorption responses 5,16,17,18 . A multiphoton technique based on non-linear optical pump-probe spectroscopy was developed, where two ultrafast laser pulses (pump and probe) were used to query the transient excited-state and ground-state photodynamics of the melanins. As a result, microscopic distribution of eumelanin and pheomelanin was directly identified in fixed and label-free human skin pigmented lesions 19,20 . However, because laser light phase modulation is critical in non-linear optical based imaging 21 , there is potential for the two endmembers of a single pump-probe wavelength combination to completely cancel each other 19 , affecting the measurement sensitivity. CARS microscopy has been used to examine cytoplasmic melanins; this involves a four-wave non-linear mixing process matching the beat frequency between the interacting pump and stokes beams with the frequency of the Raman active molecular vibration 22 . As a result, the CARS signal of the melanin molecule is enhanced compared to its spontaneous Raman scattering signal 18,22 .
Wang and colleagues (2016) successfully applied this label-free vibrational imaging to directly visualize the distribution of pheomelanin in mouse skin and hair in vivo and human skin amelanotic melanoma in situ 15 . However, the vibrational peaks from eumelanins were identified within the fingerprint region (FR) of the Raman spectrum (400 cm-1 to 1500 cm-1 frequency range). This made it difficult to isolate the signals from eumelanins because of the spectral interference from other endogenous biochemical species 15 . Taken together, the limitations of the above imaging techniques added gravitas to our decision to apply 2PM FLIM and spectral phasor plot techniques as the preferred approach in this study.

Supplementary Note 10
Inclusion of a red-brown feather, amelanotic human eye melanomas and fetal choroid for melanin controls. The pigmentation of bird plumage is mainly contributed by melanins and carotenoids. Red pigeon feathers have been previously identified to contain pheomelanin 23,24,25 , and we propose that red-brown feathers provide a mixed ratio of eumelanin/pheomelanin. Strong pheomelanin signals have been reported for amelanotic human skin melanomas imaged using CARS microscopy 15