A Cre-dependent reporter mouse for quantitative real-time imaging of protein kinase A activity dynamics

Intracellular signaling dynamics play a crucial role in cell function. Protein kinase A (PKA) is a key signaling molecule that has diverse functions, from regulating metabolism and brain activity to guiding development and cancer progression. We previously developed an optical reporter, FLIM-AKAR, that allows for quantitative imaging of PKA activity via fluorescence lifetime imaging microscopy and photometry. However, using viral infection or electroporation for the delivery of FLIM-AKAR is invasive and results in variable expression. Here, we developed a reporter mouse, FL-AK, which expresses FLIM-AKAR in a Cre-dependent manner from the ROSA26 locus. FL-AK provides robust and consistent expression of FLIM-AKAR over time. Functionally, the mouse line reports an increase in PKA activity in response to activation of both Gαs and Gαq-coupled receptors in brain slices. In vivo, FL-AK reports PKA phosphorylation in response to neuromodulator receptor activation. Thus, FL-AK provides a quantitative, robust, and flexible method to reveal the dynamics of PKA activity in diverse cell types.


Results
To characterize the utility of the homozygous FLIM-AKAR flox/flox (FL-AK) knock-in mouse line, we crossed FL-AK mice to homozygosity with Cre lines to express FLIM-AKAR in selected cell populations in the brain.We labelled Type 1 dopamine receptor (D1R) expressing spiny projection neurons (D1R-SPNs) by first crossing FL-AK with Tg(Drd1a-Cre) mice [49][50][51] , and then crossing the resulting Tg(Drd1a-Cre); FL-AK flox/+ progeny with FL-AK flox/flox to generate Tg(Drd1a-Cre); FL-AK flox/flox mice.Similarly, we labelled excitatory neurons (and a small subset of glia) in the neocortex and hippocampus with FLIM-AKAR by crossing FL-AK mice with Emx1 IREScre mice 52 across multiple generations to produce Emx1 IREScre ;FL-AK flox/flox mice.We then imaged FLIM-AKAR responses to neuromodulator receptor activation in brain slices and in freely behaving mice.

FLIM-AKAR flox/flox mice show robust and steady expression levels over time
To achieve consistent expression of the PKA activity reporter FLIM-AKAR, we generated a knock-in mouse of Cre recombinase-dependent FLIM-AKAR in the ROSA26 locus (Fig. 1a).Here, the FLIM-AKAR gene is under the control of the cytomegalovirus early enhancer/chicken β-actin (CAG) promoter.The addition of a lox-stoplox cassette makes the reporter gene Cre-dependent, allowing for selective expression of FLIM-AKAR in specific cell types.FLIM-AKAR is a fluorescence lifetime-based, genetically encoded optical reporter of PKA activity 39 .It consists of a donor fluorophore of monomeric enhanced green fluorescent protein (meGFP) and an acceptor fluorophore of dark yellow fluorescent protein (sREACH).When PKA phosphorylates the substrate consensus region within the linker region, the resulting phosphopeptide binds to the FHA1 phosphopeptide binding domain that is also in the linker region, thus bringing the donor and acceptor fluorophores closer together.This results in increased FRET and decreased fluorescence lifetime.This conformational change can also be reversed by phosphatases, which release the phosphopeptide from its binding domain through dephosphorylation (Fig. 1b).Thus, FLIM-AKAR is a phosphorylation substrate reporter that reports the balance between PKA and phosphatases.After crossing FL-AK mice to Cre lines to express FLIM-AKAR in selected cell types, we assessed FLIM-AKAR expression and functional responses with 2pFLIM.
Emx1 IREScre ;FLIM-AKAR flox/flox mice showed robust FLIM-AKAR expression across the cortex and in the hippocampus (Fig. 1c,d).In contrast, in Cre-/-; FLIM-AKAR flox/flox animals, there was very little green signal, demonstrating the Cre dependence of FLIM-AKAR expression (Fig. 1c,d).To observe cellular-level FLIM-AKAR expression across cell types, we collected both fluorescence intensity and lifetime data from acute brain slices using 2pFLIM.We observed reliable FLIM-AKAR expression throughout the cell in D1R-SPNs in Tg(Drd1cre);FL-AK flox/flox mice, and in CA1 pyramidal neurons of Emx1 IREScre ;FL-AK flox/flox mice (Fig. 1e).Interestingly, these neurons also displayed heterogeneity of fluorescence lifetime, indicating different levels of PKA phosphorylation between subcellular compartments and between cells (Fig. 1f).The expression level in homozygous FL-AK knock-in mice was comparable to mice where the sensor was delivered via IUE (Supplementary Fig. S1).
In order to determine whether FL-AK reporter mice show stable expression of FLIM-AKAR over time, we performed 2p imaging of acute striatal slices from Tg(Drd1a-Cre); FLIM-AKAR flox/flox mice across a range of ages (Fig. 2).In both the nuclear and cytoplasmic compartments, expression level of FLIM-AKAR was not significantly different between mice aged 0-6 weeks (n = 14 cells from 4 mice) and mice aged 7-14 weeks (n = 16 cells from 6 mice) (Fig. 2c,d; nucleus: p = 0.755, cytoplasm: p = 0.787; 2-tailed Mann-Whitney U test).Thus, FL-AK reporter mice show robust and consistent expression and can be used to determine PKA activity across ages.

FL-AK mice report PKA activity in response to neuromodulator receptor activation in acute slices
PKA is activated by G αs -coupled receptors, one of the most well-known being the D1R.Signaling through D1Rs stimulates adenylate cyclase activity, which produces cyclic AMP (cAMP), a second messenger that can activate PKA 53,54 .D1Rs are found on various cell types, and play major functional roles in spiny projection neurons (SPNs) of the striatum 11,18,[55][56][57] .SKF 81297 is a selective D1R/D5R agonist that increases signaling through the cAMP/PKA pathway 55,57 .In order to assess whether FL-AK reporter mice can respond to D1R activation, we imaged FLIM-AKAR in D1R-SPNs of acute striatal slices from Tg(Drd1a-Cre);FLIM-AKAR flox/flox mice.D1R-SPNs showed lifetime decreases in response to the D1R/D5R agonist SKF 81297 (1 µM), first in the cytoplasm and then in the nucleus, which is consistent with D1R activation beginning in the plasma membrane (Fig. 3a-c; nucleus: p = 1.30e− 8, cytoplasm: p = 2.27e− 12, Wilcoxon signed rank test).These results indicate that FL-AK reporter mice can report PKA activity increase in response to activation of a classical G αs -coupled receptor.
We subsequently determined whether FL-AK reporter mice can report elevated PKA phosphorylation in response to G αq -coupled receptor signaling.Although G αq signaling was not classically linked to PKA, we recently discovered that endogenous G αq -coupled receptors, such as muscarinic acetylcholine receptors (mAChRs), activate PKA 10 .This activation is mediated by parallel signaling through either calcium transients or PKC activation.Here, with Emx1 IREScre ;FLIM-AKAR flox/flox mice, we assessed whether reporter mice expressing FLIM-AKAR in CA1 pyramidal neurons showed functional responses to muscarinic receptor activation.Consistent with our previous data with IUE of FLIM-AKAR 10 , we found a fluorescence lifetime decrease of FLIM-AKAR in both the nuclear and cytoplasmic compartments of CA1 pyramidal neurons in response to mAChR activation (Fig. 4a-c; nucleus: p = 0.0069, cytoplasm: p = 0.00066, Wilcoxon signed rank test; baseline vs muscarine).Following muscarinic receptor activation, we directly activated adenylate cyclases through the application of forskolin (FSK).As expected, we saw an additional decrease in fluorescence lifetime, demonstrating a further increase in PKA activity (Fig. 4a-c; nucleus: p = 2.66e− 6, cytoplasm: p = 9.31e− 10, Wilcoxon signed rank test; muscarine vs muscarine + forskolin).These data indicate that FL-AK reporter mice can be used to detect changes in intracellular PKA activity in response to diverse neuromodulator inputs.

FL-AK mice respond to dopamine receptor activation in vivo
We examined whether FL-AK reporter mice are sensitive enough to detect PKA activation in vivo.We implanted an optical fiber into the dorsal striatum of Tg(Drd1a-Cre);FLIM-AKAR flox/flox mice, and monitored PKA activity in freely moving mice with a custom fluorescence lifetime photometry (FLiP) setup 45 (Fig. 5a).In response to intraperitoneal (IP) injection of the D1R/D5R agonist SKF 81297, but not saline, FLIM-AKAR in D1R-SPNs of  www.nature.com/scientificreports/ the reporter mice showed a fluorescence lifetime decrease (Fig. 5b,c; p = 0.03125, one-tailed Wilcoxon signed rank test).These are consistent with previous reports where we delivered FLIM-AKAR with adeno-associated virus 45 .These results indicate FL-AK reporter mice can report PKA activity in vivo.

Discussion
The FLIM-AKAR PKA activity reporter is a powerful optical tool that has the potential to unlock our understanding of the intracellular dynamics of PKA activity.FL-AK reporter mice facilitate the use of FLIM-AKAR in genetically defined cell populations specified by the Cre line with which the mice are crossed.Here, we show that FL-AK flox/flox mice exhibit robust Cre-dependent expression of FLIM-AKAR that is stable over time.They also show reliable functional responses to diverse neuromodulator signals including activation of both G αs -and  G αq -coupled receptors.Furthermore, FL-AK mice demonstrate a sufficient dynamic range to distinguish between different levels of PKA phosphorylation in vivo.FL-AK mice offer several advantages over surgical methods to deliver the reporter.First, the FL-AK mouse line generates robust and consistent expression over time.Although fluorescence lifetime is largely insensitive to sensor expression levels, the variable expression seen in surgical delivery methods can result in differential contribution of autofluorescence, leading to an apparent sensor expression-dependent lifetime response 58 .Thus, FL-AK reporter mice facilitate chronic imaging for transient PKA activation, comparison of basal PKA phosphorylation over time, and comparison of pooled results across multiple cells, mice, and experiments.Second, the mouse line eliminates the need for invasive surgeries like IUE or intracranial viral injection.Third, although not explored in this study, knock-in mouse lines can allow effective targeting of sparsely distributed or hard-totransfect cell types such as specific types of microglia [59][60][61] .Fourth, this mouse line simplifies multiplex imaging to study how PKA activity changes in relation to other critical signaling molecules in cellular processes 62 .It is worth noting that one disadvantage of knock-in lines is that Cre can be expressed in a progenitor cell type during development that would not express Cre once differentiated in adults.This can lead to deviation from the expected cell target.Should this occur, viral delivery of the reporter in adults could overcome this limitation.
Importantly, whereas this study focused on neuronal applications, FL-AK reporter mice have the potential to facilitate understanding the important roles of PKA dynamics in diverse tissues and body systems.For example, PKA has been studied in the context of immune modulation, cancer biology, and metabolic disorders such as obesity 27,28,33,[63][64][65][66][67][68] .The delivery of FLIM-AKAR via surgical methods can pose a significant technical barrier to studying PKA dynamics in these tissues.Crossing FL-AK with diverse Cre lines will create opportunities to study PKA dynamics throughout the body and better understand how this signal modulates many critical processes.

Knock-in mice
The floxed FLIM-AKAR reporter mouse line was generated by the Gene Targeting Transgenics Facility at Howard Hughes Medical Institute's Janelia Research Campus.FLIM-AKAR was knocked into the ROSA26 locus, which was demonstrated to produce robust expression of inserted transgenes 69,70 .CAG promoter and the woodchuck hepatitis virus posttranscriptional regulatory element (WPRE) were used for robust and ubiquitous expression.A lox-stop-lox (LSL) cassette was included between the CAG promoter and the FLIM-AKAR open reading frame to produce Cre dependence.The FLIM-AKAR insert was cloned into a ROSA26-pCAG-loxp-STOP-PGKNeoloxp-WPRE targeting vector 71 between the second loxp and the WPRE sequences, where PGKNeo stands for polyphoglycerate kinase I promoter driving the neomycin phosphotransferase gene (PGKNeo).
Aggregation method, where 8-10 embryonic stem cells were co-cultured with an 8-cell CD1 embryo, was used to produce chimeric mice.
Chimeric males were bred with CD1 female mice, and the female pups were crossed with the chimeric father to achieve a large number of pups for the homozygosity test to check for correct targeting.Genotyping of the floxed-FLIM-AKAR mice was performed by polymerase chain reaction (PCR) using the following primers: R26 wt gt Forward: CCA AAG TCG CTC TGA GTT GT.R26 wt gt Reverse: CCA GGT TAG CCT TTA AGC CT.CMV scr Reverse: CGG GCC ATT TAC CGT AAG TT.
PCR amplifies a fragment of 250 bp of the endogenous Rosa26 locus in wild type mice, and a fragment of 329 bp between the Rosa26 locus and the insert for floxed FLIM-AKAR mice.
The sperms of a correctly targeted F1 male were harvested, and in vitro fertilization of C57BL/6 females was performed by the Washington University Mouse Genetics Core.The progeny was bred with C57BL/6 mice for multiple generations to achieve strain stability and bred to homozygosity.The mice used in this manuscript were produced after 3 to 10 generations of breeding in a C57BL/6 background.

Implantation of optical fibers
For FLiP experiments, an optical fiber (Doric Lenses, MFC_200/245-0.37_4.5mm_MF1.25_FLT)was implanted as described previously 73 .Here, the dorsal striatum was targeted using stereotaxic coordinates of 1.1 mm anterior and 1.7 mm lateral from Bregma and 2.5 mm ventral from the pia.Four stainless steel screws were secured in the skull to better anchor the dental cement.

Two-photon fluorescence lifetime imaging microscopy (2pFLIM) and image analysis
2pFLIM was performed as previously described 73 except for the following.920 nm excitation wavelength was used to image FLIM-AKAR.For the calculation of lifetime, first, a double exponential curve was fitted to the lifetime histogram for the entire fields-of-view, where F(t) is the photon count at lifetime t, F 0 is the peak photon count, τ free and τ FRET are fluorescence lifetimes of donors that are free and that have undergone FRET respectively and are 2.14 ns and 0.69 ns for FLIM-AKAR.P free and P FRET are the corresponding fractions of these two species, IRF is the measured instrument response function, and F background is the background measurement.Then, average lifetime for a given region of interest (ROI) was calculated from 0.0489 ns to 11.5 ns with the following calculation: where F(t) is the photon count at a given time channel, and t is the lifetime measurement at that time channel.Intensity measurement was represented by the photon count/pixel of a given ROI.Change of fluorescence lifetime per cell at baseline was quantified as the absolute value of the difference between the average of the first three lifetime values of the baseline epoch and the minimum lifetime value of the baseline epoch.Change in lifetime due to drug treatment was quantified as the absolute value of the difference between the average of the last three lifetime values of the baseline epoch and the minimum lifetime value of the corresponding treatment epoch.
For Fig. 2, all the data were collected at an imaging power of 2.5 mW and between 20-35 μm from the surface of the slice.For Supplementary Fig. 1, data were collected at imaging powers between 0.55-2.5 mW at depths of 25-35 μm from the surface of the slice.To correct for systematic differences in intensity caused by difference in imaging power, the following relationship for two photon microscopy was used: where F is the photon count and P is the laser power used for imaging.

Fluorescence lifetime photometry (FLiP) and analysis
A FLiP setup was built and used as previously described 45,73 except for the following.Fluorescence lifetime and intensity data were collected at 1 Hz using our custom FLiP setup and acquisition software 45,73,74 .We calculated lifetime at each timepoint by first fitting a double exponential curve to the fluorescent lifetime histogram using a Gaussian IRF (Eq.1).Then, the average lifetime of the fitted curve (to infinity) was calculated.
Data were aligned to injection timepoints using synchronized video recordings through Bonsai (https:// bonsai-rx.org/).Location of fiber implant was subsequently assessed with histology.Data analysis was performed using MATLAB.

In vivo SKF81297 response experiments
Mice with an optical fiber implant were connected to a patch cord (Doric Lenses, MFP_200/220/900-0.37_1.5m_FCM-MF1.25_LAF) and placed in a round chamber to which they had been habituated previously.A camera was oriented to capture the entire recording chamber to assess behavior and capture injection time.Each trial

Figure 1 .
Figure 1.Expression of FLIM-AKAR across multiple cell types in the FLIM-AKAR flox/flox mouse.(a) Schematic of the gene targeting strategy to generate the FLIM-AKAR flox mice.(b) Schematic of how FLIM-AKAR works.FLIM-AKAR detects PKA phosphorylation via the change of FRET between a donor fluorophore and an acceptor fluorophore.When PKA phosphorylates the PKA consensus substrate region in the linker, the two fluorophores come closer together, resulting in increased FRET and decreased fluorescence lifetime.The process can be reversed by phosphatases.(c,d) Images of coronal slices showing FLIM-AKAR expression in the whole brain (c) and hippocampus (d) in Emx1 IREScre ;FLIM-AKAR flox/flox (left) and Cre-;FLIM-AKAR flox/flox (right) mice.Images for (d) are higher resolution versions of the same sections in (c).Images within the same panel have matching imaging conditions.(e,f) 2p fluorescence intensity (e) and lifetime (f) images of D1R-SPNs in the dorsal striatum (left) and pyramidal neurons in CA1 (right).

Figure 2 .
Figure 2. Stable expression of FLIM-AKAR in FL-AK reporter mouse over time.(a,b) 2p images of D1R-SPNs of the dorsal striatum in acute brain slices.(c,d) Quantification of nuclear and cytoplasmic photon count from two age ranges (0-6 weeks: n = 14 cells from 4 mice; 6-14 weeks: n = 16 cells from 6 mice).Data are represented as median with 25th and 75th percentiles (ns: not significant; p > 0.05, 2-tailed Mann-Whitney U test).

Figure 3 .
Figure 3. Functional response of D1R-SPNs to D1R activation in acute slices of Tg(Drd1a-Cre); FLIM-AKAR flox/flox mice.(a) Time-lapse heatmaps of 2pFLIM images of an example SPN in acute slices of the dorsal striatum in response to the D1R agonist SKF 81297 (1 µM).Dotted line indicates the location of nucleus.(b) Example traces of fluorescence lifetime response of the D1R-SPN shown in (a).(c) Quantification of change of fluorescence lifetime in response to SKF 81297 in D1R-SPNs (n = 43 cells from 6 mice).Data are represented as median with 25 th and 75 th percentiles (***p < 0.001, Wilcoxon signed rank test).

Figure 4 .
Figure 4. Functional response of CA1 pyramidal neurons in response to muscarinic acetylcholine receptor (mAChR) activation and adenylate cyclase activation in acute slices of Emx1 IREScre ; FLIM-AKAR flox/flox mice.(a) Time-lapse heatmaps of 2pFLIM images of an example CA1 pyramidal neuron in an acute hippocampal slice in response to the mAChR agonist muscarine (mus, 10 µM) and adenylate cyclase activator forskolin (FSK, 50 µM).Dotted line indicates the location of nucleus.(b) Example trace of fluorescence lifetime response of the pyramidal neuron shown in (a).(c) Quantification of change of fluorescence lifetime in response to mus and FSK in CA1 pyramidal neurons (n = 32 cells from 4 animals).Data are represented as median with 25th and 75th percentiles (***p < 0.001, **p < 0.01, Wilcoxon signed rank test).