The fluorescent dye DAPI is useful for its association with and consequent amplification of an ∼460 nm emission maximum upon binding to dsDNA. Labelling with higher DAPI concentrations is a technique used to reveal Pi polymers [polyphosphate (polyP)], with a red-shift to ∼520–550 nm fluorescence emission. DAPI–polyP emissions of ∼580 nm are also generated upon 415 nm excitation. Red-shifted DAPI emission has been associated with polyP and RNA and has more recently been reported with polyadenylic acid (polyA), specific inositol phosphates (IPs) and heparin. We find that amorphous calcium phosphate (ACP) also demonstrates red-shifted DAPI emission at high DAPI concentrations. This DAPI spectral shift has been attributed to DAPI–DAPI electrostatic interactions enabled by molecules with high negative charge density that increase the local DAPI concentration and favour DAPI molecular proximity, as observed by increasing the dye/phosphate ratio. Excitation of dry DAPI (∼360 nm) confirmed a red-shifted DAPI emission. Whereas enzymatic approaches to modify substrates can help define the nature of DAPI fluorescence signals, multiple approaches beyond red-shifted DAPI excitation/emission are advised before conclusions are drawn about DAPI substrate identification.
The fluorescent dye DAPI is commonly used to reveal DNA by microscopy. When DAPI binds to dsDNA, it exhibits a typical emission maximum of ∼460 nm upon excitation with ∼360 nm light. Its enhanced fluorescence upon interaction with dsDNA has been attributed to two bond types between the DAPI and DNA: hydrogen bonds between the AT base pairs and the amidino groups and electrostatic interaction between the phosphate groups of dsDNA and the two positively charged end-groups of DAPI . A wide range of complex interactions between DAPI and the minor groove in dsDNA has been proposed .
DAPI binding modes to DNA
Electrostatic interactions between the (N-H) group and phosphate were proposed by Wartell et al.  to describe an interaction between netropsin (an antibiotic with a similar structure to DAPI) and the A-T regions of dsDNA. An electrostatic interaction between dsDNA phosphate and netropsin or distamycin was suggested by Luck et al.  in the same year. The binding features of DAPI with DNA are solution specific; DAPI–DNA interaction can change from groove-binding to intercalation and can also induce polymer–dye adducts (Figure 1); DAPI is not only an intercalator, minor groove binder and major groove binder, but can exist with different bound forms, depending on the dye/polymer ratio, solvent type, temperature and other experimental conditions .
Schematic of DAPI–DNA interactions
DAPI also binds RNA
DAPI interacts with RNA , as it forms a very favourable intercalation complex with RNA at AU sites . This complex is similar to the DAPI complex with GC sequences in DNA. In addition, compared with the stable DAPI complex in the minor groove of AT sequences of DNA, there are weaker stabilizing interactions in the wide, shallow minor groove of RNA. There are interesting properties with respect to the fluorescence excitation and emission of DAPI bound to DNA compared with RNA . For instance, the emission maximum shifts to a shorter wavelength when complexed with DNA, whereas there is a red-shift when DAPI binds RNA. This and other spectral property differences may revolve around the distinct chemical characteristics, steric characteristics of minor and major grooves and electrostatic potentials in RNA and DNA.
Red-shifted DAPI fluorescence emission upon DAPI–DAPI electrostatic interactions
Cavatorta et al.  found unique quantum yields and time-resolved fluorescent lifetimes for DAPI complexed to nucleic acids of different composition. Although this implicated multiple binding complexes, a new binding form was revealed when a high dye/phosphate ratio (1:10) was employed. The new fluorescent species showed an emission maximum at 540 nm and was postulated to represent binding of DAPI molecules to sites in the polynucleotides that were in proximity to a previously bound DAPI molecule. The red-shift was lost when the background electrolyte was changed to 0.4 M KCl, as these DAPI–DAPI electrostatic interactions would be disrupted. Kubista et al.  also identified DAPI–DAPI interactions with circular dichroism (CD).
Red-shifted fluorescence emission at an increased dye/phosphate ratio has also been noted by Kapuscinski  for DAPI–polyadenylic acid (polyA). The study investigated binding to double-stranded nucleic acids and the role of dye–polymer condensation in changing fluorescence properties. A red-shift of dye emission spectra, compared with free and dsDNA-complexed dye, was reported. A concluding proposal suggested that condensed structures of DAPI with polyanions have altered dye properties, including enhanced fluorescence.
DAPI binding to inorganic polyphosphate with red-shifted emission
In 1980, Allan and Miller  reported that exposing Saccharomyces cerevisiae to a high concentration of DAPI (50 μg/ml) resulted in the staining of intracellular polyanion composed of Pi polymers [polyphosphate (polyP)]. Identification was associated with red-shifted, yellow-green (520–550 nm) fluorescence of the DAPI–polyP complex, compared with the DAPI–DNA complex, upon excitation with ∼360 nm light. DAPI–polyP identification and quantification has been refined with excitation at 415 nm and emission at 550 nm to minimize the influence of DAPI–DNA fluorescence .
Recent inquiry into the possible role of inorganic polyP as a Pi source for skeletal mineral (apatite) nucleation used multiphoton excitation to map DAPI–polyP fluorescence. Fluorescence of 50 μg/ml DAPI applied to dry-cut tissue sections was quantified at 580 nm to avoid convolution with DAPI–DNA fluorescence . Questions about the specificity of the fluorescing DAPI–polyP complex lead to an additional step of in situ polyP degradation by alkaline phosphatase, with subsequent identification of Pi with von Kossa staining and reduction in DAPI–polyP fluorescence in elasmobranch skeletons .
DAPI binding to inositol phosphates and heparin
A red-shifted DAPI fluorescence was also reported by Kolozsvari et al. , who demonstrated that inositol phosphates (IPs; particularly IP5 and IP6) and heparin fluoresce at ∼550 nm when complexed with DAPI. As the role of IP in apatite mineralization was investigated in the recent past , conclusive identification of in situ P-rich species in mineralization sites with DAPI fluorescence remains uncertain owing to the possible contribution of highly phosphorylated IP . Extraction and/or the application of kinases and phosphatases could further confirm identification of polyPs and IP by way of DAPI spectral signature shifts. Techniques such as Raman spectroscopy offer an in situ method for IP identification, without extraction .
DAPI fluorescence red-shifts when associated with amorphous calcium phosphate, but not hydroxyapatite
The effect of changing the substrate chemistry on DAPI fluorescence was demonstrated with amorphous calcium phosphate (ACP), which is another potential P-rich biological apatite precursor. ACP was observed to be an unstable precursor for hydroxyapatite (HAp)  and was identified in mineralizing fish skeletons [19–21]. Using methods similar to those employed to identify polyPs by DAPI , we found that DAPI powder and DAPI-labelled Ca–polyP particles fluoresce with the same wavelength maximum of 550 nm (Figures 2A and 2B). Exposing synthetic HAp to DAPI did not generate high-intensity fluorescence (results not shown), suggesting that HAp surface chemistry does not exhibit the same affinity for DAPI.
Spectrally shifted DAPI signatures upon binding to P-rich compounds
ACP particles demonstrated a similar, but transient, red-shifted fluorescence profile upon exposure to DAPI (Figure 2C). The increased blue-shift with time may be caused by transformation of the acidic, negatively-charged ACP particle to HAp in the aqueous medium . The resulting decrease in negative surface charge in surface chemistry may have decreased DAPI affinity. Altogether, this suggests that DAPI may also interact with the surface of negatively-charged particles, possibly through the electrostatic attraction between N-H groups and, in the case of polyP, phosphoryl groups (Figure 2D). A high concentration of phosphoryl groups in close proximity to each other, as presented by the polyP anion structure, may attract multiple DAPI molecules that experience dye–dye interactions. Consequently, a red-shifted fluorescence is also observed for DAPI–polyA  and DAPI–RNA .
Whereas red-shifted, yellow-green DAPI fluorescence is a simple technique that has been used recently for polyP identification, the discovery of a similar emission for DAPI associated with IP, ACP and heparin necessitates comparisons before and after sample processing, such as enzymatic action and/or independent analytical techniques for reliable substrate identification. For accurate polyP identification, other methods should be included, such as 31P-NMR and Raman spectroscopy, as used by Kolozsvari et al.  to confirm the absence of inorganic polyP in plant seeds. Biochemical techniques for polyP identification include its hydrolysis with exopolyphosphatase (PPX) and consequential reduction in DAPI–polyP fluorescence  and/or associated increase in Pi concentration.
P-rich compounds such as IP, Ca–polyP and ACP that have multiple phosphate groups or heparin with many sulfate groups in close proximity, may present a condensed, negatively-charged feature that attracts and locally concentrates DAPI. Red-shifted DAPI fluorescence is not due to specific substrate chemistry, but indicates the presence of a high density of negatively-charged surfaces or molecules that locally concentrate DAPI. This local, increased DAPI concentration enables DAPI–DAPI interactions and its resultant red-shifted fluorescence.
We thank Dr Mathieu Bennett (Department of Biomaterials, Max Planck Institute of Colloids and Interfaces) for helpful discussions, Professor Roberto Chica (Department of Chemistry, University of Ottawa) for sharing his fluorescence spectrophotometer, as well as for helpful discussions, Professor Dr Dr.h.c. Peter Fratzl (Department of Biomaterials, Max Planck Institute of Colloids and Interfaces) for research support and the reviewers for their helpful feedback.
This work was funded by the Volkswagen Foundation within the ‘Experiment!’ initiative, project number 87872 (to S.O.); and the German Research Foundation within the framework of the Deutsch-Israelische Projektkooperation DIP.
Inorganic Polyphosphate (polyP) Physiology: Held at Charles Darwin House, London, U.K., 7 September 2015.