Biology subjects

Matzinger, M.

Publications and source records attributed to Matzinger, M..

18 recordsLinked to original sources

bioRxiv2026

Making the Impossible Possible: Uniting TMT and DIA for Scalable Low-Input-Proteomics

Isobaric multiplexing using tandem mass tags (TMT) and data-independent acquisition (DIA) are two powerful techniques in quantitative proteomics, each offering distinct advantages. Combining these strategies provides synergistic benefits for low-input applications such as single-cell proteomics (SCP). However, integrating these strategies has remained challenging due to co-isolation interference and increased data complexity. Here, we present a systematic, identification-linked evaluation of a hybrid DIA-TMTpro 18-plex workflow on the Orbitrap Astral mass spectrometer (MS). Using a two-proteome interference model of human and yeast digests, we benchmarked quantitative accuracy and precision of TMT based DDA and DIA acquisition strategies across a range of isolation windows, injection times and input amounts. We provide READ (Reporter Ion Extractor and Annotation Director), an in-house developed Python-based tool which enables streamlined reporter ion (RI) quantification and co-isolation purity assessment. Using READ, TMT-based quantification becomes possible from DIA-focused search engines like Spectronaut or DIA-NN, which do not natively support TMT. We show that co-isolation interference drives both ratio compression and ratio inflation in TMT RI-based quantification regardless of acquisition mode. As an interference-resistant alternative, we further evaluated quantification via TMT complementary ions, which achieved near-label free quantitative accuracy at the cost of reduced plexing capacity and proteome depth. Together, these results establish practical guidelines for combining DIA and TMTpro on the Orbitrap Astral mass spectrometer and provide a foundation for extending DIA-TMT workflows to low-input applications such as SCP, where sensitivity, accuracy and throughput are critical.

biochemistry↗

bioRxiv2026

Defining Quality Control Standards for Single-Cell Proteomics by Inter-Laboratory Benchmarking

Single-cell proteomics can quantify thousands of proteins from individual mammalian cells, yet the absence of community-wide quality control limits biological interpretability. Here, the HUPO Single Cell Initiative presents the first inter-laboratory single-cell proteomics benchmarking study across seven laboratories using standardized 384-well plates acquired on Orbitrap Astral and timsTOF Ultra2 instruments. Centralized analysis across six DIA software tools revealed that software choice impacts identification depth and quantitative accuracy more than instrument vendor. Multi-layered quality control enabled the detection of cell-leakage during sorting, LC misconfiguration, column degradation and site-specific pipetting failures. Inter-lab quantitative correlations were strongest between instruments of the same vendor relative to cross-platform comparisons. Sequential correction for plate identity and well position recovered clean cell-type separation for confident downstream differential expression analysis. This study provides a data-driven quality control framework spanning plate design to batch correction for reproducible single-cell proteomics across laboratories and platforms.

biochemistry↗

bioRxiv2026

The Single Cell Proteomic blueprint, navigating instrumentation platforms, software tools and high-load libraries in neutrophils, RKO and A549 cells

Mass spectrometry-based single cell proteomics (SCP) is rapidly emerging as a powerful approach for biological research, with applications extending beyond in-vitro cancer cell lines. Recent advances make it possible to apply SCP to ex-vivo human cells from tissues such as the brain and pancreas, as well as to technically challenging immune populations such as neutrophils. However, these analyses remain more challenging and typically result in reduced proteomic coverage. To support the development of robust workflows for SCP data acquisition and analysis, we systematically evaluated multiple DIA search engines, search engine settings, the inclusion of high-load library samples in single-cell search spaces, the impact of contaminants, and the quantitative properties of identified proteins. These comparisons were performed across two major instrumentation platforms, Orbitrap Astral and timsTOF SCP, and across A549, RKO cells and neutrophils, three cell types differing in size and protein content. Our work here provides guidelines on the software parameters to use for SCP, instrument specific results and cell dependent optimizations of high-load libraries, as well as novel evaluation of the quantitative properties of proteins for single cell and low input proteomics.

biochemistry↗

bioRxiv2026

Pushing the limits of SCP: bacSCP, a proof-of-concept study to investigate heterogeneity of bacteria by single cell proteomics.

Single-cell proteomics (SCP) has emerged as a powerful approach to quantify protein expression variability at cellular resolution, yet most state-of-the-art workflows are tailored to eukaryotic cells with only one study exploring how single bacteria can be analyzed by mass spectrometry. Here, we established bacSCP, a protocol extending SCP to bacterial cells, facing analytical challenges such as the thick bacterial cell wall hampering lysis, the extremely small cell size and resultant low protein content, and the consequently relatively high level of contaminating proteins from external sources. Using this bacSCP pipeline, we quantified more than 50 bacterial proteins from single Bacillus subtilis and Escherichia coli cells. Upon heat stress, we reproducibly observed up to 8-fold upregulation of chaperones including GroEL, GroES, and ClpC for a B. subtilis {Delta}mcsB strain. Importantly, single-cell measurements revealed potential heterogeneity within the heat-stressed subpopulation, enabling interrogation of stress-response variability at the proteome level. These results demonstrate the feasibility of bacSCP and provide a foundation for studying bacterial stress adaptation and phenotypic diversity with single-cell proteomic resolution.

systems biology↗

bioRxiv2025

Unified down-stream analysis of crosslinking mass spectrometry results with pyXLMS

Crosslinking mass spectrometry has become the method of choice for the identification of protein-protein interactions and for gaining insight into the structures of proteins in vivo. However, connecting crosslink search engine results with down-stream analysis tools, and therefore gaining biological insight from crosslink identifications, has remained a manual and cumbersome step in the analysis that often requires expert bioinformatics knowledge. Here we introduce pyXLMS, a python package and public web application which aims to simplify and streamline this intermediate step, enabling researchers even without bioinformatics knowledge to conduct in-depth crosslink analyses. In its current state pyXLMS supports input from seven different crosslink search engines, as well as the mzI-dentML format of the HUPO Proteomics Standards Initiative. Down-stream analysis is facilitated by functionality that is directly available within pyXLMS such as aggregation, validation, annotation, filtering, and visualization. In addition, the data can easily be exported to more than ten supported down-stream analysis tools and formats. We demonstrate the applicability and benefits of pyXLMS by re-analyzing a publicly available crosslink dataset with a variety of different search engines and show how the same data analysis workflow can be applied using pyXLMS. pyXLMs is available via https://github.com/hgb-bin-proteomics/pyXLMS.

bioinformatics↗

bioRxiv2025

Low-resolution FAIMS for increased peptide coverage in low-load and single-cell proteomics

Field Asymmetric Ion Mobility Spectrometry (FAIMS) enhances signal to noise ratio filtering ions based on their differential mobility, making it an indispensable tool for single-cell and low-input proteomics. Here, we investigate the impact of FAIMS resolution tuning via electrode temperature modulation to improve peptide identification sensitivity. We demonstrate that lowering FAIMS resolution broadens the compensation voltage window, thereby increasing ion transmission. This "low-resolution" mode significantly improves peptide identifications from a concentration range of HeLa digests by up to 18%. A weaker but still beneficial effect in peptide identifications and enhanced sensitivity could also be reproduced for single-cell samples. This leads to proteomic fingerprint shifts, resulting in distinct populations in principal component analysis from the very same cell-type in dependence on FAIMS resolution. Moreover, the increased ion counts of runs employing "low-resolution" FAIMS improve the quantitative precision of low-load measurements. These findings offer a practical optimization strategy for FAIMS-based low-input proteomics workflows, that allow for improved results by changing a single setting within the MS method without the need of any change in hardware adoption or data analysis pipelines.

biochemistry↗

bioRxiv2025

In vivo crosslinking and effective 2D enrichment for interactome studies of the nucleosome.

Cross-linking mass spectrometry has evolved as a powerful technique to study protein-protein interactions and to provide structural information over the past decades. Low reaction efficiencies, and complex matrices lead to challenging system wide crosslink analysis. In this study, we improved and streamlined an Azide-A-DSBSO based in vivo crosslinking workflow employing two orthogonal effective enrichment steps: Affinity enrichment and size exclusion chromatography (SEC). Combined, they allow an effective pulling of DSBSO containing peptides and remove the background of linear as well as mono-linked peptides. We found that the analysis of a single SEC fraction is effective to yield [~]90% of all crosslinks, which is important whenever measurement time is limited, and sample throughput is crucial. Our workflow resulted in more than 5000 crosslinks from K562 cells and generated a comprehensive PPI network from whole cells as well as nuclear extracts. From 393 PPI found within the nucleus, 56 have not yet been reported in the STING database and are representing a valuable resource for investigating new molecular mechanisms and provide complementary data for future studies. We further show, that by applying DSBSO to nuclear extracts we yield more crosslinks on lower abundant proteins and showcase this on the DEAD-box RNA helicase DDX39B which is predominantly expressed in the nucleus. Our data indicates that DDX39B is present in monomeric and dimeric form together with DDX39A within the nuclear extracts analyzed.

biochemistry↗

bioRxiv2024

A journey towards developing a new cleavable crosslinker reagent for in-cell crosslinking

Crosslinking mass spectrometry (XL-MS) is a powerful technology that recently emerged as an essential complementary tool for elucidating protein structures and mapping interactions within a protein network. Crosslinkers which are amenable to post-linking backbone cleavage simplify peptide identification, aid in 3D structure determination and enable system-wide studies of protein-protein interactions (PPIs) in cellular environments. However, state-of-the-art cleavable linkers are fraught with practical limitations, including extensive evaluation of fragmentation energies and fragmentation behaviour of the crosslinker backbone. We herein introduce DiSPASO as a lysine-selective, MS-cleavable cross-linker with an alkyne handle for affinity enrichment. DiSPASO was designed and developed for efficient cell membrane permeability and crosslinking while securing low cellular perturbation. We tested DiSPASO employing three different copper-based enrichment strategies using model systems with increasing complexity (Cas9-Halo, purified ribosomes, live cells). Fluorescence microscopy in-cell crosslinking experiments revealed a rapid uptake of DiSPASO into HEK 293 cells within 5 minutes. While DiSPASO represents progress in cellular PPI analysis, its limitations and low crosslinking yield in cellular environments require careful optimisation of the crosslinker design, highlighting the complexity of developing effective XL-MS tools and the importance of continuous innovation in accurately mapping PPI networks within dynamic cellular environments. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=60 SRC="FIGDIR/small/621843v2_ufig1.gif" ALT="Figure 1"> View larger version (13K): org.highwire.dtl.DTLVardef@1821ed4org.highwire.dtl.DTLVardef@1b6180dorg.highwire.dtl.DTLVardef@1e50aaborg.highwire.dtl.DTLVardef@1f6185_HPS_FORMAT_FIGEXP M_FIG C_FIG

biochemistry↗

bioRxiv2024

Proteome-wide non-cleavable crosslink identification with MS Annika 3.0 reveals the structure of the C. elegans Box C/D complex

AbstractThe field of crosslinking mass spectrometry has seen substantial advancements over the past decades, enabling the structural analysis of proteins and protein-complexes and serving as a powerful tool in protein-protein interaction studies. However, data analysis of large non-cleavable crosslink studies is still a mostly unsolved problem due to its n-squared complexity. We here introduce a novel algorithm for the identification of non-cleavable crosslinks implemented in our crosslinking search engine MS Annika that is based on sparse matrix multiplication and allows for proteome-wide searches on commodity hardware. Application of this new algorithm enabled us to employ a proteome-wide search of C. elegans nuclei samples, where we were able to uncover previously unknown protein interactions and conclude a comprehensive structural analysis that provides a detailed view of the Box C/D complex, enhancing our understanding of its assembly and functional dynamics. Our findings provide valuable insights into the intricate regulation of cellular homeostasis and immune responses, which are conserved across species, including humans. Moreover, our algorithm will enable researchers to conduct similar studies that were previously unfeasible.

bioinformatics↗

bioRxiv2024

Challenging the Astral mass analyzer - going beyond 5200 proteins per single-cell at unseen quantitative accuracy to study cellular heterogeneity.

A detailed proteome map is crucial for understanding molecular pathways and protein functions. Despite significant advancements in sample preparation, instrumentation, and data analysis, single-cell proteomics is currently limited by proteomic depth and quantitative performance. We combine a zero dead-end volume chromatographic column running at high throughput with the Thermo Scientific Orbitrap Astral mass spectrometer running in DIA mode. We demonstrate unprecedented depth of proteome coverage as well as accuracy and precision for quantification of ultra-low input amounts. Using a tailored library, we identify up to 7400 protein groups from as little as 250 pg HeLa at a throughput of 50 samples per day (SPD). We benchmark multiple data analysis strategies, estimate their influence on FDR and show that FDR on protein level can easily be maintained at 1 %. Using a two-proteome mix, we check for optimal parameters of quantification and show that fold change differences of 2 can still be successfully determined at single-cell level inputs. Eventually, we apply our workflow to A549 cells yielding a proteome coverage of up to 5300 protein groups from a single cell, which allows the observation of heterogeneity in a cellular population and studying dependencies between cell size and cell-cycle phase. Additionally, our work-flow enables us to distinguish between in vitro analogs of two human blastocyst lineages: naive human pluripotent stem cells (epiblast) and trophectoderm (TE)-like cells. Gene Ontology analysis of enriched proteins in TE-like cells harmoniously aligns with transcriptomic data, indicating that single-cell proteomics possesses the capability to identify biologically relevant differences between these two lineages within the blastocyst.

biochemistry↗

bioRxiv2023

A method for deep and quantitative protein profiling of urine sediment, soluble and exosome fractions for biomarker research

Urine collection is painless and offers the potential for kidney liquid-biopsy(1), which appears particularly appealing with regard to the diagnosis of kidney disease (2) and patient follow-up after renal transplantation (3). From a nephrological point of view, urinary sediment and the soluble and exosome fractions of urine constitute different biological entities. We here describe a method that allows deep profiling of the protein content of the above-mentioned three fractions of urine by quantitative data-independent label-free proteomics. The method was evaluated using 19 urine samples from the Nephrology outpatient clinic at Vienna General Hospital, comprising a diverse set of chronic kidney disease (CKD) as well as patients after kidney transplantation (NTX). Peptide separation was accomplished through 60 min active gradients. A timsTOF Pro2 mass spectrometer was operated in DIA mode. The total analysis time per urine sample (three fractions) was around four hours. We demonstrate adequate technical and experimental reproducibility. Our data suggest that the protein information content of these three fractions is diverse, strengthening the importance of separate analysis. The depth of our quantitative proteomics approach permitted a detection of proteins characteristic for different parts of the nephron, such as Podocin, CD2-AP and Podocalyxin (Podocytes), SLC22A8 and SLC22A13 (proximal tubule) and Aquaporin-2 (collecting duct), suggesting that our method is sensitive enough to detect and quantify biologically relevant proteins that might serve as potential biomarkers. To the best of our knowledge, the ability to quantify up to 4000 protein groups per urine sample and more than 6000 protein groups in total makes our strategy the deepest proteome profiling study of urine to date. In conclusion, we established a method with promising figures of merit that we consider broadly applicable and useful for future clinical biomarker research studies in urine.

biochemistry↗

bioRxiv2023

A high-sensitivity low-nanoflow LC-MS configuration for high-throughput sample-limited proteomics

This study demonstrates how optimized liquid chromatography-mass spectrometry (LC-MS) conditions combined with a 50 {micro}m internal diameter (I.D.) column operated at 100 nL/min enables high-sensitivity and high-throughput sample-limited proteomics analysis, including single-cell proteomics (SCP). Through systematic parameter evaluation, an optimized method was developed, capable of identifying [~]3,000 protein groups from 250 pg of HeLa protein digest using a 10-min gradient in the direct injection workflow using data-independent acquisition (DIA) from a library-free search method. Further improvements reduced the cycle time from 20 to 14.4 min by employing a trap-and-elute workflow, permitting 100 runs/day with 70% MS utilization. A proof of principle study indicated that ca. 1,700 protein groups were identified from single-cell samples without a library for label-free quantification (LFQ). In conclusion, we demonstrate a high-sensitivity LC-MS/MS configuration that serves the needs for limited sample analysis, permitting 100 runs/day throughout.

biochemistry↗

bioRxiv2022

Robust and easy-to-use one pot workflow for label free single cell proteomics

The analysis of ultra-low input samples or even individual cells is essential to answering a multitude of biomedical questions, but current proteomic workflows are limited in their sensitivity and reproducibility. Here we report a comprehensive workflow that includes optimized strategies for all steps from cell lysis to data analysis. Thanks to convenient to handle 1 L sample volume and standardized 384 well plates the workflow is easy for even novice users to implement. At the same time, it can be performed semi-automatized using the CellenONE(R), which allows for highest reproducibility. To achieve high throughput, ultrashort gradient lengths down to 5 min were tested using advanced -pillar columns. Data-dependent acquisition (DDA), wide-window acquisition (WWA) and data-independent acquisition (DIA), and commonly used advanced data-analysis algorithms were benchmarked. Using DDA, 1790 proteins covering a dynamic range of four orders of magnitude were identified in a single cell. Using DIA, proteome coverage increased to more than 2200 proteins identified from single cell level input in a 20-min active gradient. The workflow enabled differentiation of two cell lines, demonstrating its suitability to cellular heterogeneity determination.

biochemistry↗

bioRxiv2022

Wide Window Acquisition and AI-based data analysis to reach deep proteome coverage for a wide sample range, including single cell proteomic inputs

A comprehensive proteome map is essential to elucidate molecular pathways and protein functions. Although great improvements in sample preparation, instrumentation and data analysis already yielded impressive results, current studies suffer from a limited proteomic depth and dynamic range therefore lacking low abundant or highly hydrophobic proteins. Here, we combine and benchmark advanced micro pillar array columns ({micro}PAC) operated at nanoflow with Wide Window Acquisition (WWA) and the AI-based CHIMERYS search engine for data analysis to maximize chromatographic separation power, sensitivity and proteome coverage. Our data shows that {micro}PAC columns clearly outperform classical packed bed columns boosting peptide IDs by up to 140%. Already at classical narrow isolation widths CHIMERYS boosted ID rates by a factor of 2.6 compared to the conventional search engine MS Amanda 2.0. By combining CHIMERYS with WWA, even a 4.6-fold increase in ID rates could be achieved. Using our optimized workflow, we were further able to identify more than 10,000 proteins from a single 2 h gradient shotgun analysis. We further investigated the applicability of WWA for single cell inputs and found that the choice of the optimal isolation window width depends on sample input and complexity. Using a short 5.5 cm column and very high flow rates during loading and column equilibration we improved sample throughput to [~]100 samples per day while maintaining high protein ID numbers. We believe that this is especially important for the single cell field where throughput is one of the most limiting factors. Finally, we applied our optimized workflow on immunoprecipitations of Smarca5/SNF2H and found 32 additional interaction partners compared to the original workflow utilizing a packed bed column. These additional interaction partners include previously described interaction partners of Smarca5 like Baz2b as well as undescribed interactors including Arid1a, which is also involved in chromatin remodeling and has been described as key player in neurodevelopmental and malignant disorders.

biochemistry↗

bioRxiv2021

Analysis of 8000 proteins and reduced carry over significantly increase the throughput of single-shot proteomics

In the field of LC-MS based proteomics, increases in sampling depth and proteome coverage have mainly been accomplished by rapid advances in mass spectrometer technology. The comprehensiveness and quality of data that can be generated do however also depend on the performance provided by nano liquid chromatography (nanoLC) separations. Proper selection of reversed-phase separation columns can be important to provide the MS instrument with peptides at the highest possible concentration and separated at the highest possible resolution. In the current contribution, we evaluate the use of prototype generation 2 PAC nanoLC columns which use C18 functionalized superficially porous micro pillars as a stationary phase. When comparing to traditionally used fully porous silica stationary phases, more precursors could be characterized when performing single shot data-dependent LC-MS/MS analyses of a human cell line tryptic digest. Up to 30% more protein groups and 60% more unique peptides were identified for short gradients (10 min) and limited sample amounts (10-100 ng of cell lysate digest). With LC-MS gradient times of 10, 60, 120 and 180 min, we respectively identified 2252, 6513, 7382 and 8174 protein groups with 25, 500, 1000 and 2000 ng of sample loaded on column. Reduction of sample carry over to the next run (up to 2 to 3%) and decreased levels of methionine oxidation (up to 3-fold) were identified as additional figures of merit. When analyzing a disuccinimidyl dibutyric urea (DSBU) crosslinked synthetic library, 29 to 59 more unique crosslinked peptides could be identified at a experimentally validated false discovery rate (FDR) of 1-2%.

biochemistry↗

bioRxiv2021

A molecular network of conserved factors keeps ribosomes dormant in the egg

Ribosomes are produced in large quantities during oogenesis and stored in the egg. However, the egg and early embryo are translationally repressed. Using mass-spectrometry and cryo-EM analyses of ribosomes isolated from zebrafish and Xenopus eggs and embryos, we provide molecular evidence that ribosomes transition from a dormant to an active state during the first hours of embryogenesis. Dormant ribosomes are associated with four conserved factors that form two modules and occupy functionally important sites of the ribosome: a Habp4-eEF2 module that stabilizes ribosome levels and a Dap1b/Dapl1-eIF5a module that represses translation. Dap1b/Dapl1 is a newly discovered translational inhibitor that stably inserts into the polypeptide exit tunnel. Thus, a developmentally programmed, conserved ribosome state plays a key role in ribosome storage and translational repression in the egg.

biochemistry↗

bioRxiv2021

Mimicked synthetic ribosomal protein complex for benchmarking cross-linking mass spectrometry workflows

The field of cross-linking mass spectrometry has matured to a frequently used tool for the investigation of protein structures as well as interactome studies up to a system wide level. The growing community generated a broad spectrum of applications, linker types, acquisition strategies and specialized data analysis tools, which makes it challenging, especially for newcomers, to decide for an appropriate analysis workflow. Therefore, we here present a large and flexible synthetic peptide library as reliable instrument to benchmark crosslinkers with different reactive sites as well as acquisition techniques and data analysis algorithms. Additionally, we provide a tool, IMP-X-FDR, that calculates the real, experimentally validated, FDR, compares results across search engine platforms and analyses crosslink properties in an automated manner. The library was used with the reagents DSSO, DSBU, CDI, ADH, DHSO and azide-a-DSBSO and data were analysed using the algorithms MeroX, MS Annika, XlinkX, pLink 2, MaxLynx and xiSearch. We thereby show that the correct algorithm and search setting choice is highly important to improve ID rate and FDR in combination with software and sample-complexity specific score cut-offs. When analysing DSSO data with MS Annika, we reach high identification rates of up to [~]70 % of the theoretical maximum (i.e. 700 unique lysine-lysine cross-links) while maintaining a low real FDR of < 3 % at cross-link level and with high reproducibility, representatively showing that our test system delivers valuable and statistically solid results. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=114 SRC="FIGDIR/small/465295v2_ufig1.gif" ALT="Figure 1"> View larger version (30K): org.highwire.dtl.DTLVardef@146c751org.highwire.dtl.DTLVardef@1c7a6e1org.highwire.dtl.DTLVardef@9d2ee7org.highwire.dtl.DTLVardef@1725693_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2019

Fast and highly efficient affinity enrichment of Azide-A-DSBSO cross-linked peptides.

Cross-linking mass spectrometry is an increasingly used, powerful technique to study protein-protein interactions or to provide structural information. Due to sub-stochiometric reaction efficiencies, cross-linked peptides are usually low abundant. This results in challenging data evaluation and the need for an effective enrichment. Here we describe an improved, easy to implement, one-step method to enrich azide-tagged, acid-cleavable disuccinimidyl bis-sulfoxide (DSBSO) cross-linked peptides using dibenzocyclooctyne (DBCO) coupled Sepharose(R) beads. We probed this method using recombinant Cas9 and E. coli ribosome. For Cas9, the number of detectable cross-links was increased from ~100 before enrichment to 580 cross-links after enrichment. To mimic a cellular lysate, E. coli ribosome was spiked into a tryptic HEK background at a ratio of 1:2 - 1:100. The number of detectable unique cross-links maintained high at ~100. The estimated enrichment efficiency was improved by factor 4 -5 (based on XL numbers) compared to enrichment via biotin and streptavidin. We were still able to detect cross-links from 0.25 g cross-linked E. coli ribosome in a background of 100 g tryptic HEK peptides, indicating a high enrichment sensitivity. In contrast to conventional enrichment techniques, like SEC, the time needed for preparation and MS measurement is significantly reduced. This robust, fast and selective enrichment method for azide-tagged linkers will contribute to map protein-protein interactions, investigate protein architectures in more depth and help to understand complex biological processes.

biochemistry↗