Imaging of formaldehyde in plants with a ratiometric fluorescent probe

We demonstrate that the ratiometric fluorescence monitoring of formaldehyde in live plant tissues is achieved with a newly developed ratiometric fluorescent probe, FAP, which effectively eliminated interference from other comparative analytes.


Introduction
Formaldehyde (FA), the simplest of the aldehydes with the formula H-CHO, has been widely used as an important precursor to many industrial materials and chemical compounds. 1 On the other hand, it has also been classied as a highly toxic compound to living organisms, because its strong interactions with proteins, nucleic acids and other biomolecules will lead to the inactivation of their biological activities. 2 For instance, excess exposure of exogenous FA to the human body will lead to a signicant health threat and cause many diseases, [3][4][5][6] including various cancers, heart disorders, diabetes, etc. Indeed, it has been regarded as a suspected carcinogen and well-known human toxicant. In plants, FA pollution is a widespread problem that can affect plant growth and development, causing a signicant decrease of the productivity and quality in agriculture. 7 But, interestingly, intracellular FA can be endogenously formed by normal physiological processes and spontaneously removed by complex metabolism using detoxication mechanisms. It has been demonstrated that glutathionedependent FA dehydrogenase (FALDH), as a key enzyme of FA metabolism, plays an essential role in the uptake and detoxication of FA in Arabidopsis plants. 8 Thus, the concentration of FA in living organisms is generally maintained in a dynamic state of equilibrium ranging from low to high millimolar levels. [8][9][10][11] For instance, a small pool of FA (0.1-10 mmol g À1 fresh weight) was identied in the intermediate of the C1 metabolism of plants. 12,13 The primary sources of FA in plants are caused by the processes of methanol oxidation, dissociation of 5,10methylene-THF, glyoxylate decarboxylation and other oxidative demethylation reactions. 14 Exogenous FA can also be absorbed and incorporated into the metabolism of plants. 7,15,16 As a result, phytoremediation has been extensively applied for the removal of FA from the polluted environment. However, the biological roles of FA and the mechanism of FA metabolism in plants have not yet been clearly dened, 14 and they are still challenging to study due to the lack of tools that can map the distribution of FA in live tissues. Thus, in order to understand the biological and pathological roles of FA well, it is crucial to develop sensitive methods or chemical tools for tracing and identifying FA in environmental samples and live plant tissues.
Coupled with uorescence microscopy techniques, uorescent imaging using small molecular uorescent probes to study biological species in living bio-systems has attracted great interest owing to its high spatiotemporal resolution, noninvasiveness, high sensitivity, and excellent selectivity, [17][18][19][20][21][22][23][24][25][26][27][28] which can offer the opportunity to monitor FA directly and easily in different biological contexts. However, the effective and, in particular, selective mapping of FA activity in live tissues with uorescent probes still remains a great challenge. Recently, only a few uorescent turn-on probes have been reported for monitoring FA. [17][18][19][20][21]29 Some could be successfully used for imaging FA in living cells, others were primarily applied for the in vitro determination of FA in water. These intensity-based uorescent probes may have a number of limitations imposed by the uorescent turn-on response at a single detection window, which means it is difficult to obtain accurate information when testing the FA concentration. Moreover, these turn-on responses can also vary depending on the experimental conditions, including the probe concentration, excitation intensity, temperature, and instrument efficiency. 30 To minimize most of the aforementioned articial and environmental factors, an attractive approach for the determination of FA in environmental samples and live tissues is developing ratiometric uorescent probes, due to their quantitative tracking ability and self-calibration effect. 23,31,32 So far, there is only one reported ratiometric uorescent probe for monitoring FA, 33 but its relatively low excitation in the UV range and its short emission wavelength might easily be overwhelmed by the tissue's auto-uorescence background, thereby limiting the probe's further application in live tissue. In addition, uorescent probes that have the ability to image inside live plant tissues to provide relevant information on endogenous FA are particularly attractive, but no such probe has been reported to date. Therefore, developing novel ratiometric probes that can be applied for the in vivo imaging of FA with highly sensitive and selective responses is critically important and there is strong demand. We herein present the development and application of a FA probe, FAP (Scheme 1), a new ratiometric uorescent sensor that can selectively report formaldehyde in real water samples and enable tracking of the endogenous FA activity in live tissues of Arabidopsis thaliana for the rst time.

Results and discussion
To design this new FA ratiometric uorescent probe, we focus on the 2-aza-Cope rearrangement reaction, 34 as illustrated in Scheme 1, where the starting imine intermediate can be formed by the selective capture of FA with a homoallylamino group, and it will further undergo a simple rearrangement and hydrolysis process to yield an aldehyde product, 17,18 potentially resulting in obvious changes of the uorescence properties. Thus, we envisioned that the homoallylamino group can possibly serve as the FA selective response site. Subsequently, in order to obtain a ratiometric response, we chose a well-known N,Ndimethylquinolin-6-amine as the uorescent chromophore because of its exhibited excellent photophysical and chemical properties, 35,36 which has been used in the design of other reported ratiometric uorescent probes as a uorescence reporter. We anticipated that the free probe would exhibit a relatively short emission wavelength owing to the internal charge transfer (ICT) effect between the N,N-dimethyl group and the 2-position homoallylamine of the chromophore. Upon treatment with FA, however, the homoallylamino group would change into an aldehyde group via specically activated processes, which would then cause an increase in the electron withdrawing ability, thereby inducing the red shi of the uorescence emission spectrum of the probe and producing a ratiometric response for FA. Based on these design considerations, we then synthesized the uorescent probe, FAP, in three steps starting from commercially available N1,N1dimethylbenzene-1,4-diamine (Scheme 2), and the structures of the target compounds were fully characterized by 1 H and 13 C NMR spectroscopy and mass spectrometry (see the ESI †).
As expected, the homoallylamino modication on FAP leads to relatively short absorbance and uorescence. The free FAP (20 mM) exhibited two absorption peaks at 254 nm and 353 nm, and one broad emission peak at 495 nm. However, the incubation of 20 mM FAP with 1 mM FA resulted in a remarkable red shi of the absorbance signals (23 nm and 76 nm, respectively). As shown in Fig. 1a, two obvious absorption bands with peaks at 277 nm and 429 nm newly emerged, and the intensities of the original peaks at 254 nm and 353 nm simultaneously decreased; meanwhile, accompanying the reaction of FAP with FA, the emission wavelength was also obviously shied (75 nm), affording an important enhancement of the uorescence emission band at 570 nm and a signicant decrease at 495 nm (Fig. 1b). These in vitro responses of FAP validate the previous design strategy and conrm that FAP is a good sensor for in vitro formaldehyde determination. In order to elaborate the reaction mechanism, we then monitored the reactions between FAP and FA using HPLC and analysed the reactions using MS and IR spectroscopy. The reactions with different incubation times showed clean conversion, and a new peak (m/z ¼ 201.1), iden-tied as the product with the expected mass of the aromatic aldehyde (3), began to appear aer 1 h of incubation ( Fig. S1 and S2 †). The reactions of different FA concentrations also showed a gradual decrease in the probe peak and a corresponding increase in the product peak with increasing FA concentration (Fig. S2 †). In addition, a signal of the freshly formed -CHOgroup was also clearly observed in the IR spectrum aer treatment with FA (Fig. S3 †). Taken together, these ndings show that the formation of the aldehyde is indeed promoted by reacting with FA.
To assess the response time of FAP towards FA, a solution of FAP (20 mM) was added to FA (2 mM) in PBS buffer (10 mM, pH 7.4, containing 1% MeCN) for different incubation times and then measured by a uorescence spectrometer (Fig. S4 †). Importantly, a signicant ratiometric change was observed even aer 5 min, accompanying a pronounced enhancement of the uorescence intensity at 570 nm and a distinct reduction at 495 nm, whereas the emission band at 570 nm was gradually enhanced with the extension of the incubation time even aer 180 min. Specically, the uorescence ratio I 570 /I 495 increased by 9-fold upon reaction with FA for 120 min. In order to observe the relatively obvious uorescence changes at an appropriate incubation time, we chose a reaction time of 120 min to further investigate the pH effect on the probe, FAP. We measured the uorescence signal ratios of I 570 /I 495 towards FA (2 mM) in PBS buffer with various pHs ranging from 3.0 to 11.0 and, surprisingly, found that FAP shows a relatively wide range of pH application (Fig. S5 †). Importantly, the remarkable ratiometric uorescence response and the wide range of pH application of FAP are advantageous to minimizing the interference of the uorescence background and increasing the delity of the detection signal.
Encouraged by these results, we sought to investigate the sensitivity of FAP for FA, and we assessed the uorescence response of FAP (20 mM) with different concentrations of FA (0-3 mM) in PBS buffer at 37 C. The incubation of FAP with increasing concentrations of FA resulted in a gradual increase of the emission intensity at 570 nm and a corresponding decrease at 495 nm ( Fig. 2 and S6 †). Strikingly, an excellent linear dependence (R 2 ¼ 0.98827) between the uorescence signal ratios of I 570 /I 495 and the concentrations of FA (0-200 mM) was observed. Moreover, to further evaluate the validity and stability of the slope, different concentrations of FA (1.5, 4.5 and 9.5 equivalent) were spiked into the PBS buffer and incubated with FAP (20 mM) for 120 min (Fig. S7 †). Aer a calculation using a calibration curve, seen in Fig. 2b, the concentration of formaldehyde was determined, and the resultant FA recoveries ranged from 98% to 106%, revealing that FAP has potential capability for quantitatively monitoring FA. The limit of detection (LOD) of FAP for the detection of FA was 0.5 mM in aqueous solution based on the signal-to-noise ratio (S/N ¼ 3).
The selectivity of the uorescent probe is an important parameter for assessing if it is suitable for biological application. To determine whether FAP could detect FA specically, we then examined the uorescence signal and compared the ratio values of the uorescence intensity at 570 nm to 495 nm (I 570 / I 495 ) aer incubation with a series of comparative analytes, including benzaldehyde, methylglyoxal, oxalaldehyde, GSH, H 2 O 2 , amino acids, etc. As seen in Fig. 2c and S8, † the largest change of the uorescence signal ratio was observed upon the addition of FA, while the addition of 100 equiv. of the other analytes displayed a negligible response. FAP displayed a 9-13fold greater response toward FA than the other analytes (Fig. 2c). Additionally, exposing FAP to a mixture of FA and other species still yielded a signicantly increased uorescence ratio (I 570 /I 495 ) (Fig. S9 †). These results clearly indicate that FAP exhibits excellent selectivity for FA over other analytes. Collectively, these selective and sensitive assays unambiguously suggest that FAP has potential capability for tracking FA in environmental and biological samples.
Having established FAP determination, we next moved on to evaluate the practicability of FAP. We initially carried out experiments using real-water samples from the Changjiang River, as well as so drink and rice water samples. FA (2 mM) was previously spiked into these real-water samples and incubated with FAP (20 mM) for 2 hours (Fig. 2d). Obvious changes of the uorescence signal ratio were observed aer incubation of FA in these buffer systems, while the so drink and rice water samples only showed relatively moderate changes that might be because FA was partially consumed via the interaction with biomolecules. But fortunately, these changes of the uorescence signal ratio were enough to help us achieve the determination of FA in the real-water samples. Moreover, we prepared dried FA test papers by dipping paper into a FAP solution and further evaporating to remove the solvent (Fig. 2d inset). Interestingly, using this simple FA test paper to detect FA solutions with various concentrations exhibited quick color changes from green to yellow within several seconds, indicating the potential utility of FAP as a simple sensor for tracking FA in environmental samples.
To address the capability of FAP for mapping FA in live tissue, we employed FAP to image FA in the live root tip tissue of Arabidopsis thaliana using a confocal uorescence microscope. As shown in Fig. 3, the root tip tissues of Arabidopsis thaliana were pretreated with FAP (50 mM) for 45 min, and subsequently different concentrations of exogenous FA (0 or 3 mM) were added and incubated for another 120 min. Gratifyingly, FAP can effectively penetrate the cell wall and cell membrane, and can display a bright uorescence in the green channel and relatively weak uorescence in the red channel, suggesting that live Arabidopsis thaliana tissues have a low level of intracellular FA activity. 12,13 Importantly, upon incubation with a typical stress concentration of exogenous FA (3 mM), a signicant change of the uorescence signal ratio is observed in the ratiometric uorescence image (F red /F green ) (Fig. 3). This indicates that the alteration of the intracellular FA levels in live plant tissues can be reected by the uorescence change of FAP. Notably, these FA concentrations fall well within the scope of physiological concentration, treating the condition of formaldehyde stress in plants, and are estimated to be 0.1-10 mmol g À1 in the fresh weight of a plant, 13 and up to 2-8 mM in the external treatment of HCHO stress. 14,15,37 To further investigate whether FAP can be applied to detect the change of endogenous FA levels in live plant tissues, we pretreated live root tip tissues of Arabidopsis thaliana with 10% MeOH stress for 1 h to induce the production of endogenous FA, 38 then incubated with FAP (50 mM) for another 3 hours and imaged aer washing with PBS (Fig. 4). The endogenous FA freshly formed via MeOH pretreatment results in an obvious change of the uorescence signal ratio in the ratiometric uorescence image compared with the negative control. Furthermore, the stimulation of endogenous FA uptake with 1 mM ophenanthroline, a known inhibitor of formaldehyde dehydrogenase, 39 also yielded a statistically signicant increase in the uorescence ratio (Fig. 4). All of these data suggest that FAP is capable of ratiometric monitoring of endogenous FA activity in Arabidopsis thaliana in vivo. Finally, we tried to apply FAP to measure the endogenous formaldehyde concentration in the fresh root and leaf tissues of Arabidopsis thaliana. Spiked HCHO was used as the internal standard in the prepared tissue homogenate. The spiked samples were subsequently incubated with the 20 mM probe for 2 hours and then monitored (Fig. S10 †). We found that the concentrations of endogenous formaldehyde in the fresh homogenates of the Arabidopsis thaliana root and leaf tissues are 1.8 mmol g À1 and 4.2 mmol g À1 , respectively. This result is consistent with the previous report that the average endogenous FA concentration was estimated to be 0.1-10 mmol g À1 of the fresh weight of the plant. 13 Taken together, these ndings demonstrate that FAP displays the ability of detection of the biologically relevant endogenous formaldehyde in live plant tissues.

Conclusions
In summary, we have successfully developed a ratiometric uorescent probe, FAP, for monitoring formaldehyde based on the N,N-dimethylquinolin-6-amine uorophore for the rst time. This probe displayed a signicant red shi (75 nm) in the emission proles and an obvious change of the uorescence signal ratio for the response of FA. We have demonstrated that this highly selective and effective probe has potential capability for monitoring FA in real-water samples and live plant tissues. In vivo imaging results show that FAP can not only effectively map exogenous FA, but also can sensitively track endogenous FA. To the best of our knowledge, this is the rst example of a uorescent probe-based methodology for the ratiometric imaging of endogenous FA in plant tissues. With these insights, Fig. 3 Confocal microscopy images of Arabidopsis thaliana incubated with FAP (50 mM) for 45 min at 37 C (a-d), followed by the addition of 3 mM HCHO (e-h) and incubation for another 2 hours. (a and e) The green channel (475-520 nm), (b and f) the red channel (545-595 nm), (c and g) the merge images of the bright-light field, green channel and red channel, and (d and h) the ratio images of the red to green channel. l ex ¼ 405 nm, scale bar ¼ 50 mm. (i) Quantification of the imaging data (d and h). n ¼ 3, error bars were AESD. Statistical analyses performed with a twotailed Student's t-test with unequal variance, ***p value < 0.0001. Fig. 4 Confocal microscopy images of endogenous FA in Arabidopsis thaliana via pretreatment with or without 10% MeOH and 1 mM ophenanthroline in PBS buffer for 1 h, followed by exchange into fresh PBS and further incubation with FAP (50 mM) for 3 h at 37 C before imaging. (a, e, and i) Green channel (475-520 nm), (b, f, and j) red channel (545-595 nm), and (c, g, and k) merge images of the bright-light field, green channel and red channel, and (d, h, l) the ratio images of the red to green channel. l ex ¼ 405 nm, Scale bar ¼ 50 mm. (m) Quantification of the imaging data (d, h, l). n ¼ 3, the error bars were AESD. Statistical analyses performed with a two-tailed Student's t-test with unequal variance, ***p value < 0.001.
we anticipate that the further application of FAP to the uorescent imaging of FA in other plants will enable better understanding of its metabolism mechanism, thus promoting the progress of FA related physiological and pathological studies.